*Article* **Microbiome Analysis of the Rhizosphere from Wilt Infected Pomegranate Reveals Complex Adaptations in Fusarium—A Preliminary Study**

**Anupam J. Das 1,2, Renuka Ravinath 1, Talambedu Usha 2, Biligi Sampgod Rohith 3, Hemavathy Ekambaram 4, Mothukapalli Krishnareddy Prasannakumar 5, Nijalingappa Ramesh <sup>1</sup> and Sushil Kumar Middha 4,\***


**Abstract:** Wilt disease affecting pomegranate crops results in rapid soil-nutrient depletion, reduced or complete loss in yield, and crop destruction. There are limited studies on the phytopathogen *Fusarium oxysporum* prevalence and associated genomic information with respect to *Fusarium* wilt in pomegranate. In this study, soil samples from the rhizosphere of different pomegranate plants showing early stage symptoms of wilt infection to an advanced stage were collected from an orchard situated in Karnataka, India. A whole metagenome sequencing approach was employed to gain insights into the adaptations of the causative pathogen *F. oxysporum*. Physicochemical results showed a drop in the pH levels, N, Fe, and Mn, and increase in electrical conductivity, B, Zn, Cl, Cu was observed in the early and intermediate stage samples. Comparative abundance analysis of the experimental samples ESI and ISI revealed an abundance of Proteobacteria phyla *Achromobacter* sp. 2789STDY5608625, *Achromobacter* sp. K91, and *Achromobacter aegrifaciens* and Eukaryota namely *Aspergillus arachidicola*, *Aspergillus candidus,* and *Aspergillus campestris.* Functional pathway predictions implied carbohydrate binding to be significant (*p* < 0.05) among the three experimental samples. Microbiological examination and whole microbiome analysis confirmed the prevalence of *F. oxysporum* in the soil samples. Variant analysis of *F. oxysporum* revealed multiple mutations in the 3IPD gene with high impact effects. 3-Isopropylmalate dehydratase and carbohydrate-active enzymes could be good targets for the development of antifungals that could aid in biocontrol of *F. oxysporum*. The present study demonstrates the capabilities of the whole metagenome sequencing approach for rapid identification of potential key players of wilt disease pathogenesis wherein the symptomatology is complex.

**Keywords:** microbiomics; soil metagenomics; DNA sequencing; wilt; rot; *Punica granatum*

155

#### **1. Introduction**

Pomegranate is a widely cultivated fruit crop with its origins traced to Turkey and Iran. The crop is extensively cultivated in various parts of India and India has emerged as the leading producer of pomegranate globally [1–3]. There are extensive reports on the medicinal properties of the pomegranate viz it's antimicrobial [4], antihyperglycemic [5], anticancerous [6,7], its nutraceutical [8], pharmaceutical [9], and cosmeceutical [10] applications due to the presence of a wide range of nutrients, secondary metabolites such as alkaloids and flavonoids [2]. There are also reports of anti-inflammatory properties and the

**Citation:** Das, A.J.; Ravinath, R.; Usha, T.; Rohith, B.S.; Ekambaram, H.; Prasannakumar, M.K.; Ramesh, N.; Middha, S.K. Microbiome Analysis of the Rhizosphere from Wilt Infected Pomegranate Reveals Complex Adaptations in Fusarium—A Preliminary Study. *Agriculture* **2021**, *11*, 831. https://doi.org/10.3390/ agriculture11090831

Academic Editors: Alessandra Durazzo and Anna Andolfi

Received: 19 June 2021 Accepted: 18 August 2021 Published: 30 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

potential of the pomegranate juice and peel against various disorders [2,3,11–13] and protection from UV photodamage [14]. However, cultivating the crop has been challenging due to its susceptibility to diseases and pest infestations which results in a drastic reduction in the yield and quality of the fruit. Major diseases are caused as a result of bacterial and fungal infections. Some of the fungal pathogens reported are *Colletotrichum acutatum* [15] *Trichoderma* spp., *Botrytis cinerea, Aspergillus niger, Penicillium* spp., *Alternaria* spp., *Colletotrichum gloeosporioides, Pestalotia brevista,* and *Pilidiella granati* [16–19] anthracnose disease of the flower. The infection results in the abortion of the flower leading to a reduction in the yield [15]. The manifestations of infection caused by *Pilidiella granati* are crown rot, twig blight, and dieback with common symptoms of necrosis in fruits and twigs respectively during the early stage of infection [15,19]. Among the bacterial diseases, blight disease is one of the serious challenges faced by farmers in India. The causative organism has been identified as *Xanthomonas campestris*. pv. *punicae*. Yield loss of up to 80% has been reported in Bangalore, Karnataka as a result of an epidemic outbreak. The pathogen infects the entire plant. Epidemic outbreaks are also reported in Andhra Pradesh, Maharashtra, and Delhi [20]. The detection of the fungal and bacterial pathogens is generally done through the isolation of the organism followed by culturing them. The identification is done based on the morphological characters and physiological and biochemical tests. These methods are highly labor-intensive, time-consuming, need expertise [21] and only the cultivable organisms can be identified. These limitations were later overcome by PCR based diagnostic method. The identification and detection of *P. granati* were carried out through the nested PCR method. Species-specific primers were designed and the method could effectively detect the pathogen in the fruits of pomegranate [22]. In another study, the phytopathogen *Xanthomonas campestris* pv. *punicae*, causing blight disease in pomegranate was detected using ERIC-PCR-Generated genomic fingerprints. A relationship was established between the fingerprints and virulence pattern of the blight-causing pathogen [20]. These methods have limitations with respect to specificity as they are not based on DNA sequencing.

16S rRNA gene sequencing is an excellent approach to reveal the identity of the pathogen as they are signature specific sequences in bacterial species with higher accuracy. Bacterial wilt disease in *Cucurbita maxima* in China caused by *Ralstonia solanacearum* was identified by 16S rRNA gene sequencing of the isolates obtained from the plants infected with wilt. Pathogenicity analysis revealed that all the isolates belonged to *Ralstonia solanacearum* [23]. Investigations on determining the microbiota associated with symptomatic and non-symptomatic bacterial wilt-diseased banana plants were also done using 16S rRNA metagenome sequencing. Illumina MiSeq platform was used for sequencing. The results revealed the predominance of *Ralstonia* in the pseudostem of the symptomatic diseased plant compared to non-symptomatic [24]. The findings could also throw light on the role of endophytic microbes revealed through sequencing studies in conferring tolerance to the disease. Many successful studies have been carried out in fruit crops and vegetables, where 16S rRNA gene sequencing has emerged as an excellent tool for the detection of the associated plant pathogens. Another newly reported disease in pomegranate is the Bacterial root-bark necrosis disease and wilt in pomegranate, which was found to affect the plant entirely. A recent study by Ajaysree and Borkar, 2018, shed light on the symptomatology of the disease that includes symptoms of wilt disease on the leaves and stem such as yellowing of leaves, followed by leaf fall and wilting of branches. The study reports complete death of the plant with no recovery in a period of 2–3 months. On the other hand, reports suggest that the roots of plants show symptoms of root-bark necrosis. 16S rRNA sequencing facilitated the identification of the pathogenic bacterium *Klebsiella pneumoniae* [25].

To explore the correlation between monocropping followed in banana and the Fusarium wilt incidence, the soil samples from such fields were subjected to sequencing of 16S rRNA genes for bacteria and internal transcribed spacer using the MiSeq platform for fungal identification. The findings led to the conclusion that monocropping significantly

increased the incidence of Fusarium wilt [26], with the help of 16S metagenomics the role of the cropping system in disease management.

16S rRNA sequencing using the 454 platforms could accurately reveal the bacteria associated with the nematodes infesting pine trees. This association is responsible for the wilt disease of pine. 25 Operational Taxonomic Units could be analyzed based on 97% of similarity in the sequences of the library. The microbial diversity revealed Alphaproteobacteria, Betaproteobacteria, Gammaproteobacteria, and Bacteroidetes [27]. These findings are vital for adopting the proper control measures for wilt disease as it is influenced by the nematodes as well as the associated microbiota establishing a unique ecosystem.

However, despite all the benefits of employing the 16S metagenomics approach, there are certain limitations to consider prior to planning a soil metagenomics study. Firstly, soil microbial diversity is vast, and exploring the soil communities with a targeted approach that considers quantifying relative abundances of taxa may remain incomplete in terms of its functional potential. Secondly, the resolution is dependent on the databases employed. There are large scale efforts put towards developing database and tools to improve classification of bacterial communities and their diversity [28–30] The emergence of long read platforms have offered potential solutions to help sequence the entire 16S rRNA using the Nanopore or PacBio platforms. Some studies have reported higher microbial identification and taxonomic resolution as compared to the short amplicon sequencing despite the higher depth from platforms such as Illumina [31]. To specifically identify fungi, the gene cluster within the 18S ribosomal RNA is considered and the repetitive internal transcribed spacer (ITS) sequences are used. Furthermore, 18S rRNA sequencing comes with its limitations particularly with extraction methods showing biased results [32], primer-biases in PCR resulting in amplification of certain taxa preferentially [33], the copy numbers of the small subunit (SSU) rRNA genes [34], sequencing errors [35,36] and remnant DNA amplification [37].

The whole metagenome sequencing approach has helped address some of these limitations pertaining to targeted metagenome approaches. Microbe-pest-host associations are complex and their adaptations remain elusive. The whole metagenome approach is a powerful method to not only screen microbes but also facilitate understanding of plantmicrobe-soil interactions and the disease pathogenesis in plants. The advanced Illumina Novaseq 6000 (Illumina) offers a unique possibility to perform soil microbial characterization [38,39]. Despite the large data outputs from these platforms, bioinformatics analysis of the data employing server and cloud-based analytics services have enhanced the speed and efficiency of analysis [40]. There are limited studies on the phytopathogen *Fusarium oxysporum* prevalence and associated genomic information with respect to Fusarium wilt in pomegranate. In the present study, we demonstrate the implementation of shotgun sequencing using the whole metagenome approach to study the pathogenomics of wilt disease in *Punica granatum* caused by Fusarium, wherein the symptomatology is complex.

#### **2. Materials and Methods**

The present study involved screening the physiochemical parameters of the soil samples from the rhizosphere of the infected plants. The total microbial counts were estimated. Employing conventional microbiological methods the soil samples were plated on specialized media to confirm the presence of the pathogen *Fusarium oxysporum* and *Aspergillus niger*. Following which genomic DNA was isolated from soil samples and the quality control of the samples was performed. Whole metagenome shotgun sequencing was carried out. Thereafter, the data was subject to bioinformatics analysis to estimate the relative abundances of the microbes, and the functional predictions were performed. Finally, variant analysis was carried out to screening for possible targets that could provide key leads to understand the adaptations of the pathogen. An overview of the methodology adapted in the study is provided in Figure 1.


**Figure 1.** Overview of the experimental workflow from sample collection to identifying key players in microbial adaptation. The figure depicts the wet laboratory and dry laboratory methods employed in the present study.

#### *2.1. Site Description and Sampling*

Soil samples were collected from a pomegranate orchard close to Chikkaballapur (13.3907◦ N, 77.6880◦ E) from Karnataka, India. The orchard has been used for cultivating the crop over a span of 5 years and the farmer suffered huge losses due to the reducing fruit yield attributed to severe pest infestations. In the past year alone, the farmer suffered a loss in fruit yield by over 36%. The land was surveyed and post-harvest, without the application of any pesticides or antibacterial or antifungal agents, samples were collected in December 2019 from the rhizosphere of 5 plants from each category showing similar symptoms. The symptoms were categorized as early signs of infection, moderate signs of infection, on and severe infections, in triplicates and pooled (Figure 1). The plants were identified as early-stage infection (ESI) with early symptoms of wilt, intermediate stage of infection (ISI), and advanced stage of infection (ASI) on the basis of physical examination of the leaves, stem, fruits, and roots (Figure 2).

**Figure 2.** Physical symptoms of the plants, figures depict the physical symptoms of the infected plants—ESI; Early-stage infection, ISI; Intermediate stage infection, and ASI; Advanced stage infection (**a**–**d**). Other wilt-associated symptoms such as yellowing of leaves and root knots observed (**e**), rotting of fruit (**g**), complete defoliation (**f**), brown decay, and sporulation (**h**) observed are depicted.

#### *2.2. Physicochemical Characterization and Total Microbial Count Estimation*

All the physical and chemical characterizations were carried out based on the procedures provided by [41]. pH values, electrical conductivity, were estimated by the electrometric method [41]. The total microbial counts were estimated using the protocols provided in IS 5402 and IS 5403 for the total bacterial and total fungal count respectively [42,43]. Each reading was collected in duplicates.

#### *2.3. Isolation of Fusarium oxysporum, Aspergillus niger from Soil Samples*

39 g of potato dextrose agar powder (catalog no. M096, HiMedia) was added in 1 L sterile water and it was thoroughly mixed. The media was autoclaved at 15 psi pressure at 121 ◦C for 15 min. Test Samples (1 mL) were 10-fold diluted in 9 mL of water (10<sup>−</sup>1). From that sample was serially diluted up to (10−2) and (10−3). All three dilutions were plated on Selective media by spread plate technique. The plates were incubated in both aerobic chambers at 37 ◦C for 24–48 h for bacteria and 27 ◦C for 48–72 h. After 24–48 h incubation colonies were observed and recorded [42] (Tables S1 and S2).

#### *2.4. DNA Extraction and Quality Control*

DNA extraction was carried out based on the protocol by Amorim et al. [44]. Nanodrop was used initially to test the purity of DNA (OD260/OD280) (NanoDrop, Wilmington, DE, USA). Agarose Gel Electrophoresis was performed to assess DNA degradation and potential contaminations (Figure S1) and finally, Qubit 2.0 was used to quantify the DNA concentration precisely.

#### *2.5. Library Construction and Quality Control*

Qualified DNA was cut into fragments by the restriction enzyme. The construction of the DNA libraries is through the processes of end repairing, adding A to tails, purification, PCR amplification, and Libraries were sequenced by Illumina high-throughput sequencer with paired-end sequencing strategy. The libraries, that passed the QC, were then fed into sequencers after pooling according to their effective concentration and expected data volume.

#### *2.6. Whole Meta-Genome Sequencing*

The qualified libraries are fed into sequencer Illumina Novaseq 6000 (sequencing facility of Novogene Co. Ltd., Beijing, China) after pooling according to its effective concentration and expected data volume. The detailed protocol is provided in the supplementary data, Table S3.

#### *2.7. Data Analysis*

Raw Data QC of individual samples was conducted using FastQC (parameters: default) (https://www.bioinformatics.babraham.ac.uk/projects/fastqc/; accessed on 10 January 2020, Table S4). FastQ Screen v0.14.0 (https://www.bioinformatics.babraham.ac.uk/ projects/fastq\_screen/; accessed on 11 January 2020) was employed to screen the host genome sequences (References: GCF\_007655135.1, GCA\_002837095.1, GCA\_002864125.1, GCA\_002201585.1) from the raw data (parameters: –tag –filter '00000'; configured with bowtie2-2.3.5.1 and default Adapters). Host screened data was further validated using fastq-pair v1.0 and then assembled using metaSPAdes v3.13.0 (parameters: default) [45], metagenomic classification and visualizations using Kraken2 (parameters: –use-names –paired –gzip-compressed; database: built on 12 November 2020) [46] and pavian (https: //github.com/fbreitwieser/pavian/; accessed on 10 January 2020) respectively. SPAdes assembled genome was subjected to gene prediction using MetaGeneMark (parameters: -f 3 -a -d -k -v -m MetaGeneMark\_v1.mod) [47]. The predicted nucleotides were searched against NCBI NR database (Downloaded in March 2020) using Diamond v0.9.30 (parameters: -k 1) BlastX. Annotations were further meganized (parameters: default) using MEGAN v6 (relevant database were Downloaded in March 2020) [48].

Genome resolved metagenomics of Individual samples was performed using the SqueezeMeta pipeline v1.3.0 in sequential mode with MegaHIT assembler (default parameters). Short contigs (<200 bps) were removed and contig statistics were estimated using prinseq. RNAs were predicted using Barrnap. 16S rRNA sequences were taxonomically classified using the RDP classifier. tRNA/tmRNA sequences were predicted using Aragorn [49]. ORFs were predicted using Prodigal. Similarity searches for GenBank, eggNOG, KEGG, were done using Diamond HMM homology searches were done by HMMER3 [50] for the Pfam database. Read mapping against contigs was performed using Bowtie2. Binning was done using MaxBin2 [51] and Metabat2 [52]. A combination of binning results was performed using the DAS Tool [53]. Bin statistics were computed using CheckM. Bins with at least 50% completeness and <20% contamination were selected and subjected to annotation using enrichM's annotation module (https://github.com/geronimp/enrichM, accessed on 20 May 2021) against KO, PFAM, EC, and CAZY databases. The generated annotation matrices were subjected to enrichment using enrichM's enrich module. Alternatively, functional analysis of the metagenomic samples was also done using MG-RAST online server (https://www.mg-rast.org/; accessed on 22 January 2020) with default parameters.

#### *2.8. Prediction of Protein Functions*

Processed FASTQ reads of individual samples were used to search most popular databases that include protein databases, protein databases with functional hierarchy information, and ribosomal RNA databases namely RefSeq, IMG, TrEMBL, Subsystems, KEGG, GenBank, SwissProt, PATRIC, eggNOC, KO, GO, COG, RDP, LSU, SSU and NOG as a part of the MG-RAST analysis.

#### *2.9. Variant Analysis*

Based on the analysis of the processed reads and the resultant assemblies obtained from metaspades, the organisms *Fusarium oxysporum* and *Aspergillus niger* were used from the samples ESI and ISI respectively. BWA [53] was utilized for indexing (parameters: *index*) and mapping (parameters: mem) the pre-processed reads to *A. niger CBS 513.88* (NCBI Accession: GCF\_000002855.3), *and F. oxysporum* f. sp. *lycopersici* 4287 (NCBI Accession: GCF\_000149955.1) reference genomes. Then the aligned reads were converted to bam and sorted using Samtools v.1.6 [54] (parameters: *sort -l 9*) and duplicate reads were marked using GATK MarkDuplicates v.4.1.9.0 [55], followed by adding and replacing the read groups using GATK AddOrReplaceReadGroups. The reference dictionary was created for the genome using Samtools (parameters: *dict)* and the reference genome was also indexed using Samtools (parameters: *faidx*) to aid the variant calling process. GATK HaplotypeCaller (parameters: *-ERC GVCF -GQB 50*) and GATK GenotypeVCFs to generate known variants for base quality score recalibration using GATK BaseRecalibrator, followed by GATK ApplyBQSR. Furthermore, the recalibrated alignments were again run through GATK HaplotypeCaller (parameters: *-ERC GVCF -GQB 50*) and GATK GenotypeVCFs to identify the variants, followed by selecting identified SNPs and INDELs separately using GATK SelectVariants (parameters: *–select-type-to-include SNP)* (Parameters: *–select-type-toinclude INDEL*) and extracted SNPs were further masked using GATK VariantFiltration (parameters: *-mask -mask-extension 5 –mask-name* "*INDEL*") to tag the SNPs identified in and around INDELs. Finally, the masked SNPs and filtered INDELs were merged using GATK MergeVcfs [56]. In order to predict the variant effects, reference genomes were prepared using snpEff v.5.0d [57] (parameters: *build-genbank*) and variant summary and effect predictions were obtained.

#### *2.10. Statistical Analysis*

Output from MEGAN in. SPF format was used for statistical analysis and was performed using STAMP 2.1.3 (http://kiwi.cs.dal.ca/Software/STAMP; accessed on 10 January 2020, [58]). The one-sided G-test (w/Yates' + Fisher's) with asymptomatic confidence intervals (0.95) using the Benjamini–Hochberg FDR method was implemented [59].

#### **3. Results**

#### *3.1. Physical Examination*

Physical examination of the pomegranate plants, with respect to their, roots, leaves, stems, and fruits showed symptoms of wilt were considered for the study. Examination of the fruits showed the presence of rot disease was deemed for study.

Considering the ESI plant sample, few leaves showed mild yellowing (Figure 2b,e) and on examination of the roots, root knots were observed (Figure 2e). A few of the fruits showed black spots with mild discoloration. In the ISI, the dark brown coloration of the stem was observed, the fruits were darkly colored irregular spots with cracking (Figure 2f) and leaves showed yellowing, presence of moist, dark-colored irregular spots (Figure 2b). The ASI sample showed complete defoliation (Figure 2c) fruits that had completely turned dry with dark-brown pigmentation (Figure 2h), the root systems were dry and reduced with elongated galls, and dark brown coloration of the stem which has turned completely dry was observed (Figure 2f). ASI sample had complete yield reduction with no recovery (Figure 2).

#### *3.2. Physiochemical Properties*

The pH of the ESI sample was found to be 7.73 and electrical conductivity was estimated to be 135 μs/cm. Macronutrient and micronutrient analysis of ESI revealed 0.20% of total nitrogen (N), 0.0084% Phosphorous (P), 0.011% Potassium (K), 0.92% organic Carbon (C), 15 ppm Chloride(Cl), 0.98% Iron (Fe), 9.5 ppm Manganese (Mn), 26.9 ppm Copper (Cu), 24.8 ppm Zinc (Zn), and 3.4 ppm Boron (B). The pH and electrical conductivity of the ISI sample were estimated to be 6.35 and 139 μs/cm, respectively. In ISI samples the total N was calculated to be 0.19%, followed by P (0.010%), K (0.011%), C (0.93%), Cl (18 ppm), Fe (0.93%), Mn (9.1 ppm), Cu (29.4 ppm), Zn (30.9 ppm), B (4.1 ppm). The pH of the ASI sample was found to be 6.63 and electrical conductivity was estimated to be 180 μs/cm, followed by total N (0.20%), P (0.011%), K (0.014%), C (0.97%), Cl (21 ppm), Fe (0.98%), Mn (9.6 ppm), Cu (31.4 ppm), Zn (33.2 ppm), B (4.3 ppm). The pH was found to be altered in the ISI sample (6.35) and ASI sample (6.63) as compared to that of the ESI sample (7.73). Furthermore, reduced levels of N, Fe, and Mn micronutrients were reported in the ISI sample as compared to the ESI sample. Whereas, the micronutrients B, Zn, Cl, Cu, were found to be higher in the ISI sample when compared to the ESI sample (Tables 1 and S5).

**Table 1.** Physicochemical Characteristics and total microbial count of the soil samples.


EC—Electrical conductivity.

#### *3.3. Total Microbial Counts*

The total bacterial counts and total fungal were estimated to be 1968 g/cfu, 2240 g/cfu and 2126 g/cfu for the ESI, ISI, and ASI, respectively. The total bacterial counts in the ISI sample showed a significant increase as compared to ESI and ASI. Similarly, the total fungal counts showed a significant increase in the ISI sample, which was estimated to be 170 g/cfu, as compared to ESI (154 g/cfu) and ASI (154 g/cfu) (Tables 1 and S5).

#### *3.4. Sequence Information*

13 GB of high-quality raw data per sample for ESI, ISI, and ASI was generated using the Illumina Novaseq 6000 sequencer. The complete protocol information and sequencing results are tabulated in Tables 2 and S4. The BioProject Id is PRJNA701747. The BioSample Ids for the samples ESI, ISI, and ASI are SAMN17910186, SAMN17910187, and SAMN17910188, respectively. The rarefaction curve of the experimental data sets ESI, ISI and ASI are provided in Figure S2.



#### *3.5. Microbial Abundance Analysis*

Table 3 shows an overview of the taxonomic hits distribution in all three samples of ESI, ISI, and ASI. The ESI sample showed early symptoms of wilt disease. Employing the MG-Rast server, authors obtained maximum reads that mapped to Bacteria—2,076,360 (96.54%), followed by Eukaryota—63,248 (2.94%), Archaea—8979 (0.42%), unclassified sequences 1311 (0.06%), and Viruses 935 (0.04%) at the kingdom level. Actinobacteria 1,026,641 (57.52%), Proteobacteria 450,739 (25.25%), Planctomycetes 55,347 (3.10%), Ascomycota

52,339 (2.93%), Chloroflexi 37,935 (2.13%), Bacteroidetes 33,727 (1.89%), Firmicutes 29,671 (1.66%), Verrucomicrobia 21,659 (1.21%), Acidobacteria 20,072 (1.12%), Cyanobacteria 11,919 (0.67%), unclassified (derived from Bacteria) 7962 (0.45%), Gemmatimonadetes 7030 (0.39%), Deinococcus-Thermus 4694 (0.26%) and Euryarchaeota 4231 (0.24%) were mapped at the phylum level.

**Table 3.** Taxonomic hits distribution.


The ISI sample showed further symptoms of wilt disease and maximum reads from this sample also mapped to Bacteria—2,192,380 (96.08%), followed by Eukaryota—79,978 (3.50%), Archaea—7089 (0.31%), unclassified sequences—1496 (0.07%) and Viruses—939 (0.04%) at the kingdom level. At the phylum level the mapped reads revealed Actinobacteria—995,904 (52.92%), Proteobacteria—517,416 (27.49%), Planctomycetes—72,331 (3.84%), Ascomycota— 69,904 (3.71%), Chloroflexi—39,681 (2.11%), Bacteroidetes—38,024 (2.02%), Firmicutes—33,232 (1.77%), Verrucomicrobia—28,599 (1.52%), Acidobacteria—25,715 (1.37%), Cyanobacteria— 13,685 (0.73%), Gemmatimonadetes—9358 (0.50%), unclassified (derived from Bacteria)— 9248 (0.49%), Deinococcus-Thermus—4978 (0.26%), and Euryarchaeota—4290 (0.23%). The all major symptoms of wilt disease were seen in ASI sample. Our analysis revealed that the most reads were mapped to Bacteria 2,348,985 (97.61%), followed by Eukaryota 43,462 (1.81%), Archaea 10,928 (0.45%), unclassified sequences 2003 (0.08%) and Viruses 1001 (0.04%) at the kingdom level. Furthermore, at the phylum level, Actinobacteria 955,903 (49.62%), Proteobacteria 562,154 (29.18%), Planctomycetes 102,778 (5.34%), Chloroflexi 54,360 (2.82%), Firmicutes 37,538 (1.95%), Bacteroidetes 36,944 (1.92%), Ascomycota 34,814 (1.81%), Acidobacteria 33,599 (1.74%), Verrucomicrobia 33,042 (1.72%), Cyanobacteria 18,085 (0.94%), unclassified (derived from Bacteria) 11,289 (0.59%), Gemmatimonadetes 8945 (0.46%), Deinococcus-Thermus 6399 (0.33%), and Euryarchaeota 5509 (0.29%) were identified. The alpha diversity for ESI was 351, ISI was 367 and ASI was 401 species.

Comparative statistical analysis of the samples ESI and ISI revealed 35,554 features in all after filtering out unclassified reads, of which 29,747 mapped to Bacteria, 4582 to Eukaryota, 1214 to Archaea, 1 to none. From the total number of features, 79 were found to be significant with corrected *q*-value =< 0.05 (Figure 3b). Considering which of the microbes were more abundant in the ESI sample, the top differentially abundant microbes belonged to Proteobacteria phyla *Achromobacter* sp. 2789STDY5608625 with a count of 299 in ESI and 8 in ISI samples. *Achromobacter* sp. K91 from the phyla Proteobacteria followed by the *Achromobacter* sp. 2789STDY5608625 with 236 (ESI) and 4 (ISI). *Achromobacter aegrifaciens* followed the two Proteobacteria phyla members with 23. The parent sequence count 2975 (ESI) and 588 (ISI). *Microbacterium* sp. SUBG005 from phyla Actinobacteria and *Agrobacterium larrymoorei* from the Proteobacteria phyla were the subsequently most abundant bacteria (Table 4). Furthermore, comparing the ISI and ASI samples, *Streptomyces* sp. FxanaC1, *Streptomyces* sp. F12 and *Rhizobium* sp. NFACC06-2 was estimated as being the top three differentially abundant species. Amongst the Eukaryota, the *Aspergillus arachidicola* was found to be differentially abundant, followed by *Aspergillus candidus* and *Aspergillus campestris* all from the phyla Ascomycota. With the ISI and ASI comparison, *Aspergillus nomius* and *Aspergillus ochraceoroseus* were found to be species that were prevalent in the ISI sample and significantly lower in the ASI sample (Tables 4 and S6).

**Figure 3.** Microbial diversity and abundance. (**a**) The top-ranked microbes at the phylum level are depicted along with a comparison between ESI (blue) and ISI (purple); (**b**) and, ISI and ASI (pink); (**c**) depict the top-ranked microbes and the proportions and difference between the proportions with 95% confidence intervals along with the *q*-value (corrected) (*q* value < 0.001).


**Table 4.** Relative Abundance of Microbial species.

The Figure 3 depicts an overview of the kingdoms and corresponding microbes. The heatmap in the outer circle represents the prevalence among the three experimental samples as per the scale provided (a). The top-ranked microbes at the phylum level are depicted along with a comparison between ESI (blue) and ISI (purple) (b) and, ISI and ASI (pink) (c) depict the top-ranked microbes and the proportions and difference between the proportions with 95% confidence intervals along with the *q*-value (corrected) (*q* value < 0.001).

#### *3.6. Pathway Predictions*

The output files from MEGAN were used to search the InterPro2GO, a resource of protein information [60], under the Molecular functions category. The top hits included catalytic activity—oxidoreductase activity, transferase activity, and hydrolase activity. The other major functions were transporter activity, ion binding, and nucleotide binding. Under biological process, the top hits were mapped to Metabolic processes, transport, DNA, and RNA metabolic processes. Under cellular component, the intrinsic component of membrane and membrane functions were highlighted (Figure 4). eggNOC, a database for functional annotations, orthology, and gene evolution [61] revealed the major pathways related to metabolism, information storage, and processing, cellular processing, and signaling. The major pathways in metabolism included amino acid metabolism and transport, energy production and conversion, and carbohydrate metabolism and transport, inorganic ion and transport, lipid transport and metabolism, secondary metabolites biosynthesis, transport and catabolism, coenzyme transport and metabolism, and nucleotide transport and metabolism. In information storage and processing, the major pathways were transcription, replication, recombination and repair, and translation, ribosomal structure, and biogenesis. Under the cellular processes and signaling category, the pathways were signal transduction and mechanisms, cell wall/membrane/envelope biogenesis, Post-translational modifications, protein turnover, chaperones, and defense mechanisms. SEED functional annotation using SEED [62] highlighted Metabolism and stress response, defense, and virulence. Under metabolism, the predictions showed fatty acids, lipids, and isoprenoids mainly indicating acyl carrier protein (Figure S3). Statistical analysis of the predicted pathways revealed significant hits obtained from the InterPro2GO database. Carbohydrate binding was found to be significant (*p* < 0.05) between ESI and ISI. Kyoto Encyclopedia of Genes and Genomes (KEGG) [63] hits included K03088; RNA polymerase sigma-70 factor ECF sub-family, K12132; Eukaryotic-like serine/threonine protein kinase and K01990; ABC-2 type transport system ATP-binding protein (Tables 5, S7 and S8).

**Figure 4.** Pathway hits from InterPro2GO, databases. The top-ranked pathways are represented as per the color scheme three samples ESI (blue), ISI (purple), and ASI (pink) (**a**). The top-ranked pathways are represented (**a**). A comparison between ESI (blue) and ISI (purple) (**b**) and, Significant hits represented show the top-ranked pathways along with the proportions and difference between the proportions with 95% confidence intervals considering a *q*-value (corrected) (*p* < 0.05).

**Table 5.** Pathway predictions. The top hits from the most popular databases have been furnished in the table below.


\* Information storage and processing; +Cellular processes and signaling.

#### *3.7. Genome Resolved Metagenomics*

Following confirmation with microbiological methods, normalized read counts obtained for *F. oxysporum* in the ESI were found to be 209, in ISI it was 120 and 94 in ASI. Using the SqueezMeta pipeline, a total of 13, 8, and 5 genome bins were recovered from samples ISI, ASI, and ESI respectively. Based on CheckM analysis, 4, 4, 2 bins from samples ISI, ASI, and ESI were found to be at least 50% complete and with less than 15%. Significantly enriched terms (Mann–Whitney U test; *p* value < 0.05) of these bins against KO, PFAM, and other databases are shown in Table 5 and complete enrichment results are provided as Supplementary Figures S4–S6; Tables S7 and S8.

#### *3.8. Variant Analysis*

*F*. *oxysporum* genome in the sample ESI was analyzed to screening for variants using the GCF\_000149955.1 as reference (Tables S9 and S10). The variant analysis showed 81 SNPs out of which multiple mutations were observed in the 3-Isopropylmalate dehydratase (IPMD) predicted as high impact effects, which is required for fungal pathogenicity [66]. IPMD is encoded by LEU2 and involved in the leucine biosynthetic pathway.

Variant Analysis of *A. niger* genome from the ISI sample with reference (GCF\_00000285 5.3) showed three genes mutated with high impact names histidine kinase J7, MFS transporter, and NADH-ubiquinone oxidoreductase subunit. While two of the genes, histidine kinase and the MFS transporter gene, showed multiple mutations with high impact variations namely frameshift mutations, NADH-ubiquinone oxidoreductase subunit gene showed one variation leading to a frameshift mutation.

ANI\_1\_1000064|histidine kinase J7 showed an insertion at position 1764099 of A/ACG AGT. The other high-impact variant was a deletion at 1764101 GTCCTT/G. Predictions revealed three high-impact variants in the ANI\_1\_2008144|MFS transporter gene. The first one was an insertion at 1799089 position A/ACGCGCTTC, the second one was again an insertion at position 1799092 G/GTGCGT and the third one was an insertion at position 1799094 C/CG. In the ANI\_1\_742164|NADH-ubiquinone oxidoreductase subunit a large insertion was found at the position 1286777 T/TCGAGAACTCGAAGTTCGGACCCTCGACG ATGGCATCGACC.

#### **4. Discussion**

Pomegranate has been used for a wide range of health benefits making it a commercially important crop. India is the leading producer of pomegranate with Maharashtra, Karnataka, Odisha, Tamil Nadu, Gujarat, Rajasthan, Chattisgarh, Telangana, and Nagaland states contributing to India's major producer of the fruit crop. Karnataka, which is the second-largest contributor of fruit produce to India, faces a number of challenges in crop management due to wilt, anthracnose, bacterial blight, and heart rot. In the present study, we explored the soil samples from an Orchard in the Chikkaballapur district of Karnataka. Wilt infection in the orchard resulted in a 36% yield loss to the farmer. A shotgun metagenomics approach was employed and the microbial communities in soil samples were screened.

Comparative analysis of the samples ESI and ISI revealed 35,554 features in all after filtering out unclassified reads, of which the majority of the features mapped to Bacteria (29, 747), followed by Eukaryota (4582) and Archaea (1214). In our analysis, we reported 79 features to be significant (corrected *q*-value =< 0.05). The top differentially abundant microbes prevalent in the ESI sample to Proteobacteria phyla *Achromobacter* sp. 2789STDY5608625, ESI (299), and ISI (8) samples. *Achromobacter* sp. 2789STDY5608625 was followed member of the sample phyla, *Achromobacter* sp. K91, ESI (236), and ISI (4). *Achromobacter aegrifaciens* followed the two Proteobacteria phyla members with 23. *Microbacterium* sp. SUBG005 from phyla Actinobacteria and *Agrobacterium larrymoorei* from the Proteobacteria phyla were the subsequently most abundant bacteria. There have been reports of members of the genus *Achromobacter* employed as biocontrol agents against *Fusarium oxysporum* causing wilt in other plants [67,68]. The role of the microbes from this genus could be explored for their biocontrol potential against *F. oxysporum*. Furthermore, comparing the ISI and ASI samples, *Streptomyces* sp. FxanaC1, *Streptomyces* sp. F12 and *Rhizobium* sp. NFACC06-2 was estimated as being the top three differentially abundant species. Amongst the Eukaryota, the *Aspergillus arachidicola* was found to be differentially abundant, followed by *Aspergillus candidus* and *Aspergillus campestris* all from the phyla Ascomycota. With the ISI and ASI comparison, *Aspergillus nomius* and *Aspergillus ochraceoroseus* were found to be species that were prevalent in the ISI sample and significantly lower in the ASI sample.

We particularly screened *F. oxysporum* as a causative pathogen for Wilt disease in pomegranate after assessing the physical symptoms of the plant. The presence of *F. oxysporum* was confirmed both with microbial isolation and metagenomics validation. The presence of early symptoms of Wilt was reported in the ESI sample in our study. Furthermore, the fruits showed early symptoms of rot disease. The sample was screened for *A. niger* the causative pathogen for rot disease. The plant with more noticeable symptoms of the disease in our study was the ISI sample. In the rhizospheric samples of the ISI sample, we reported an abundance of *A. niger* with a decrease in *F. oxysporum*. A decline in Fusarium species with an increase in Aspergillus species was observed in the plants

from ESI and ISI samples respectively. Variant analysis of *F. oxysporum* showed multiple high-impact mutations on the IPMD gene. IPMD has been reported in other studies for its role in fungal pathogenesis. IPMD is encoded by LEU2 and involved in the catalysis of leucine biosynthesis particularly in the conversion of 3-isopropylmalate (3-IPPM) to 2-ketoisocaproate (2-KIC). Intriguingly, another important finding from this study is with respect to the carbohydrate binding pathway which is one of the significant hits. Phytopathogens are known to synthesize carbohydrate-active enzymes (CAZymes) also known as plant cell wall degrading enzymes (PCWDE) [69], which can also function as Carbohydrate binding modules (CBM). CAZymes are required for pathogenesis as well as growth [70,71]. It may be reasonable to assume that targeting IPMD and CAZymes could be a good strategy for the development of antifungals which could aid in biocontrol of *F. oxysporum*.

The present study took advantage of the current state-of-the-art sequencing platform, the Illumina Novaseq 6000 platform that provides higher resolution in screening and identification microbial communities. The approach aided in the identification of certain key targets that are linked to the pathogenicity of Fusarium. However, further research is being carried out to particularly validate the key findings of this study. In this study, we demonstrate the capabilities of the whole metagenome sequencing approach in identifying potential key players of wilt disease affecting the pomegranate plant, wherein the symptomatology is complex.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/agriculture11090831/s1, Figure S1: DNA Extraction, Figure S2: Rarefaction curves from experimental data sets ESI, ISI and ASI, Figure S3: Pathway predictions, Figure S4: KEGG Annotations, Figure S5: COG Annotations, Figure S6: PFAM Annotations. Table S1: Isolation of *Fusarium oxysporum,* Table S2: Isolation of *Aspergillus niger,* Table S3: DNA Sequencing Information, Table S4: Data Quality Summary (R2-ESI, NP-ISI and NC-ASI), Table S5: Physicochemical and Microbial Analysis of the soil samples, Table S6: Relative Abundance of Microbial species, Table S7: Pathway predictions, Table S8: PFAM Abundance Overview, Table S9: Variant Analysis, Table S10: Number variants by type.

**Author Contributions:** The idea and Funding, S.K.M.; Writing, original draft preparation, experimental work, A.J.D.; review, R.R.; analytics support, B.S.R.; supervision, N.R. and S.K.M.; funding acquisition, M.K.P. review, T.U.; wet laboratory methods support, H.E.; This work is a part of the Ph.D. work of A.J.D. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** The sequence information is available publically for this project at https://www.ncbi.nlm.nih.gov/bioproject/?term=PRJNA701747/, accessed on 10 January 2020 (Bioproject number PRJNA701747).

**Acknowledgments:** We thank the DST-FIST program level O under Govt of India and BISEP program under Govt of Karnataka at Maharani Lakshmi Ammanni College for Women for extending their generous support for this study. This manuscript was funded by University of Agricultural Sciences, Bangalore.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Perspective* **Metrology, Agriculture and Food: Literature Quantitative Analysis**

**Alessandra Durazzo 1,\*, Eliana B. Souto 2,3, Ginevra Lombardi-Boccia 1, Antonello Santini <sup>4</sup> and Massimo Lucarini 1,\***


**Abstract:** Great attention has been given in recent years to the relationships between metrology, agriculture, and food. This study aims at providing an analysis of the literature regarding the relationships between metrology, agriculture, and food. The Scopus online database has been used to extract bibliometric data throughout the search string: TITLE-ABS-KEY (Metrology\* AND Agriculture\* OR Food\*), and the VOSviewer bibliometric software was used to visualize results as bubble maps. The novelty character of this perspective paper is to indicate and point out the main research themes/lines addressing the relationships between metrology, agriculture, and food by analyzing: (i) the authors of the published papers; (ii) the type of paper; (iii) the countries and institutions where the research is developed. Bibliometrics allows one to holistically examine entire scientific areas or sub-fields to get new qualitative and quantitative insights. These results represent a useful tool for identifying emerging research directions, collaboration networks, and suggestions for more in-depth literature searches.

**Keywords:** metrology; agriculture; food; biodiversity; literature quantitative analysis

#### **1. Introduction**

Nowadays, an integrated, multidisciplinary, and interoperable approach can be seen as a modern way to research food and an innovative challenge to analyze and model agro-food systems following a holistic approach [1]. A great challenge is to identify a unique dimension of food-agricultural aspects in terms of quality and safety [2]. The recent work of Brown [3], remarks how metrology remains a unique important effort, and outlines the importance of updating the concept of metrology: it proposes a new feature— 'measuring measurement'—, emphasizing the characteristic meta-thought associated with the discipline, which distinguishes it from any routine measurement.

Metrology is the science of measurements, that is the discipline that deals with defining the procedures for performing correct measurements. The International Bureau of Weight and Measures in 2004 defined metrology as "the science of measurement, embracing both experimental and theoretical determinations at any level of uncertainty in any field of science and technology" [4].

Nonetheless, metrology is not only meters, kilograms, and atomic clocks; the science of measurement wants to be closer to everyday life, for example taking into account the "farm to fork model", reaching the consumer, certifying the origin of food. Metrology has always supported the needs of the technological world, and today it is accompanying the spread of completely new technologies. The application of metrology to the environment and food is the last frontier in this research area.

173

**Citation:** Durazzo, A.; Souto, E.B.; Lombardi-Boccia, G.; Santini, A.; Lucarini, M. Metrology, Agriculture and Food: Literature Quantitative Analysis. *Agriculture* **2021**, *11*, 889. https://doi.org/10.3390/ agriculture11090889

Academic Editor: Isabel Lara

Received: 14 July 2021 Accepted: 10 September 2021 Published: 16 September 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Today, the agri-food sector represents a strategic asset for all countries in Europe and not only there, being one of the most important socio-economic activities, and it is crucial for providing employment, the supply of healthy and high-quality food, and facilitating the integration of small and medium-sized enterprises (SMEs), representing 99% of all businesses in the European Union in the food chain [5]. The focus of consumer needs is represented by food quality and authenticity, and in this direction food traceability and safety become critical factors in ensuring the quality and protection of food consumer interests. Furthermore, providing healthy and sustainable diets is a challenge of the agri-food sector [6].

Metrology is also a tool, throughout the use of advanced data analysis methods, at the service of the so-called "precision agriculture", which combines satellite and drone images with those of sensors and actuators in order to identify, for example, the most efficient interventions in relation to the real cultivation needs and the biochemical and physical characteristics of the soil. In fact, in recent years the need to enhance the resilience of communities and territories, the reduction of the consumption of natural resources and chemical and phytosanitary fertilizers, and the optimization of agricultural production has forced the development of innovative techniques for crop management. Thus, metrology may allow, thanks to the experience in the field of meteorological forecasts, the knowledge of historical and actual agronomic data, the coordination of the activities related to the reduction of emissions and the development of information technologies for precision agriculture, such as networks of sensors, geolocation systems and agrometeorological models.

In order to achieve climate-neutrality by 2050, while also considering pandemic situations such as the COVID-19 pandemic, the challenge is to promote economic recovery and return to facing global goals, e.g., clean growth and climate change, for example by means of innovative technologies as well as the management of big data within the perspective of sharing and dissemination at the e-cloud interface [7,8]. The novelty character of the proposed perspective work is to give a current snapshot on the relationships existing between metrology, agriculture, and food, and to indicate related current directions and collaboration networks by analyzing research themes and major contributors with reference to country/regions, institutions and types of published papers. In fact, the bibliometric analysis reveals the applications of metrology principles in the agricultural and food fields.

#### **2. Materials and Methods**

The overall landscape of the literature in the research field of metrology, agriculture and food relationships has been investigated through a bibliometric analysis. The literature quantitative research analysis consisted in the following main steps: (i) literature search based on Scopus online database; (ii) data extraction and analysis.

#### *2.1. Literature Search*

In July 2021, a search for metrology, agriculture, food relationship publications was carried out based on the Scopus database content. The Scopus online database (https://www.scopus.com/home.uri, accessed on 3 July 2021) was used to extract bibliometric data using the search string: TITLE-ABS-KEY ("Metrology\*" AND "Agriculture\*" OR "Food\*"). Publications mentioning the relevant words or their derivatives in the title, abstract, or keywords were identified throughout the applied search strategy. The evaluated parameters were: trends of publications and citations, document type, authorship, country/region and institution.

#### *2.2. Data Extraction and Analysis*

Bibliographic data, e.g., publication year, publication count, citation count, document type, authorship, countries/regions and institutions, were assessed. For the basic analyses the functions of the Scopus web online platform named "Analyze" and as "Create Citation Report" were used. The "full records and cited references" were exported to VOSviewer software (version 1.6.16, www.vosviewer.com, accessed on 3 July 2021) for additional processing procedures/operations.

The terms utilized in the title abstracts of publications and keywords of publications were analyzed by the above mentioned VOSviewer software (v.1.6.16, 2020) [9–11]: the paragraphs were broken down into words and phrases, and linked to the publications' citation data, in order to visualize the results as a bubble map. This approach has been previously used in different fields of study, including food, nutraceutical, and chemical areas [12–15]. In a term map, the bubble size indicates how frequently a term is mentioned in the articles. Two bubbles that are positioned more closely to each other reflect that the terms coappeared more often in the selected publications. The average citations per publication (CPP) are given by the color of a bubble.

Five was set as the minimum number of occurrences of a keyword. Out of the 3166 keywords, 108 met the selected threshold, and 3 of them were manually excluded.

#### **3. Results**

Three hundred and twelve publications were returned by the search: they covered the time range from 1970 to 2021 and were cited collectively 2183 times. The publication and citation trends for the relationships between metrology, agriculture and food research are reported in Figure 1. The first publication is the Proceedings of the Instrument Society of America (ISA) Silver Jubilee Conference and Exhibit on advances in instrumentation, located in Philadelphia on 26–29 October 1970 and also including subjects such as metrology, analysis instrumentation and food industry [16].

**Figure 1.** Publication and citation trends for the relationships between metrology, agriculture, and food research. (Bibliometric data were extracted from the Scopus online database).

The most recent "Review" is focused on the global situation of reference materials in assuring the quality and safety of the most consumed beverages in the world, i.e., coffee, cocoa, and tea, by discussing aspects related to reference material preparation processes, as well as the results of homogeneity and stability tests and their application, together with an overview of the patents developed for food. The authors remarked in the conclusion a clear need to develop certified reference materials and reference materials for these beverages for other analytes of interest, such as chlorogenic acids and other phenolic compounds [17]. The most recent "Article" showed the application of a smart pansharpening approach using kernel-based image filtering, as an example of numerous applications of remote sensing image fusion in monitoring, metrology, and agriculture [18]. Another application of innovative technologies is given by Fiorani et al. [19] on the use of a photoacoustic

laser system for food fraud detection as a reliable technique for the rapid screening of counterfeited ingredients in the supply chain, which needs further development.

Additionally, the most cited paper (300 times) is a paper by De Chiffre et al. [20] which reported industrial applications of computed tomography in different fields including the food industry.

The distribution of the types of documents relative to the 312 publications retrieved are shown in Figure 2. "Article" accounts for 50.3%, followed by "Conference paper" (26.3%), "Review" (8.0%), and "Conference Review" (6.4%). Among "Book", it is worth mentioning the one published in 2007 by Meinrath and Schneider [21] entitled "Quality assurance for chemistry and environmental science: metrology from pH measurement to nuclear waste disposal": the basic metrological concepts for measurements in chemistry and geochemistry like traceability, ISO uncertainties or cause-and-effect diagrams were discussed, and applications of metrological techniques in highly complex situations, i.e., in thermodynamics, geochemical modeling, hydrology and radioactive waste disposal, were given. Another book focuses on the description of voluntary standards, mandatory technical regulations, conformity assessment (testing and measurement of products), certification, quality and quality management systems as well as other management systems such as environmental, social responsibility and food safety management systems [22].

**Figure 2.** Distribution of the type of document. (Bibliometric data were extracted from the Scopus online database).

A book published in 2010 entitled: "Metrology in Industry: The Key for Quality" described an analysis of the metrological requirements needed to ensure quality, along with the organization of metrology, mastering the measurement process approach, the bank of measuring instruments, the traceability to national standards, measurements and uncertainties, the measuring environment, and others [23].

Among the "Editorial" category, the editorial published by Chirico and Bonavolontà [24] entitled: "Metrology for agriculture and forestry 2019" addressed recent advances in integrated monitoring and modelling technologies for agriculture and forestry.

Figure 3 reports the most productive authors. It should be noted that 'Anon' as Anonymous was originally ranked second by the Scopus 'Analyse Results' function, and it is not listed in Figure 3. Additionally, we underline that some of the most productive authors participate in the same collaborative papers.

**Figure 3.** Most productive authors. (Bibliometric data were extracted from the Scopus online database).

Castanheira, I. (n = 8) resulted as the most productive author. The oldest work of this author regarding this matter pointed out the need to have reference materials in order to monitor the intake of nitrates [25]. Her most recent papers are represented by two conference proceedings presented at the 22nd World Congress of the International Measurement Confederation, IMEKO 2018, held from 3 to 6 September 2018 [26,27]; in one paper [26], the authors presented new Research Infrastructure METROFOOD-RI within the framework of the European Strategy Forum on Research Infrastructures (ESFRI) with the aim of promoting metrology in food and nutrition fields through the constitution of a well-organized and structured network of physical and electronic facilities: cross-cutting research activities and deliver advanced services were performed, covering several areas at the interface of different typologies of users: food business operators, research/academy, food control agencies, food policy makers, consumers/citizens. The core-services of METROFOOD-RI are the development and production of new (certified) Reference Materials (RMs). In the second one, published by Zoani et al. [27], feasibility studies for the development of new food matrix-Reference Materials (RMs) were presented. The feasibility studies included the following procedure Reference Materials preparation; procedures and guidelines definition for collecting characterization results and for processing the obtained data; Reference Materials characterization; homogeneity and stability studies; data processing and result evaluation, and three candidates for new Matrix-Reference Materials were tested: rice grains, rice flour, and lyophilized oyster tissue.

Castanheira' most highly cited paper (cited 11 times) focuses on ensuring food integrity by means of the integrated use of metrology and the application of FAIR (acronym for: findable, accessible, interoperable, and re-usable) data principles throughout the experience of the pan-European project METROFOOD-RI [28]. Among her papers, it is worth mentioning, also, the Conference Proceeding concerning the European Strategy on metrology in food composition databanks presented at the 20th IMEKO World Congress held from 9 to 14 September 2012 in Busan. In that document, the use of the International System of Units, modes of expressions and Reference Materials were indicated as being among the most important metrological tools for improving the quality of data in national Food Composition Databanks [29].

It is worth mentioning that METROFOOD-RI is composed of a Physical Infrastructure (P-RI) and an electronic infrastructure (e-RI). Among the relevant papers reported on the topic, like the ones by Zoani, C. and Zappa, G., it is worth mentioning the work of Alexandre [30], which describes the facilities which have been inventoried and classified in a database, providing an organized overview of the capacities of the distributed P-RI. These data were presented at the 3rd IMEKOFOODS Conference: "Metrology Promoting Harmonization and Standardization in Food and Nutrition", held in Thessaloniki from 1

to 4 October 2017. On the other hand, another work, presented at the same Conference by Presser et al. [31], and coauthored also by Zappa, G., described the development of a pilot e-RI where several datasets from different countries were used and interrelated to integrate national e-resources into a European-wide e-RI providing new functionalities. Additionally, a further position paper on METROFOOD-RI and its e-component were presented at the IEEE International Conference on Big Data, 2019, held in Los Angeles from 9 to 12 December 2019 [32].

The most cited paper of Iyengar, V. is a note focusing on metrological concepts for enhancing the reliability of food and nutritional measurements, including: high-quality reference standards, validated methods, robust sampling practices, proven calibration approaches, natural matrix reference materials, speciation chemistry, the assessment of measurement uncertainty and establishment of traceability links, certified reference materials to facilitate one aspect of traceability, and proficiency testing [33].

The most cited paper of Otake, T. is a research addressing the development of certified reference material—NMIJ CRM 7504—for the quantification of two pesticides in brown rice [34]. Another research from the same author was published in Food Chemistry in 2013 on the development of apple certified reference material for the quantification of organophosphorus and pyrethroid pesticides [35].

Figures 4 and 5 show the most productive countries/territories and institutions, respectively. The most productive country is represented by the United States with 48 documents. Among the documents reported for the United States, a book chapter was published on opportunities and limitations for metrology, represented by testing for foods derived from modern biotechnology [36]. An "Editorial" published by Iyengar in 2007 addressed differing perceptions of metrology in physics, chemistry, and biology [37]. Interesting results are also reported in a "Note" of Iyengar [33], previously described, and a "Note" by Koch et al. [38] on measurement science for food and drug monographs in the perspective of a global system. Among the "Article" category, the paper of Koch and Ma [39] on the approach of interfacing chemical metrology with pharmaceutical and compendial science adopted by United States Pharmacopeia indicates another relevant area of interest for the metrological approach.

**Figure 4.** Most productive countries/territories. (Bibliometric data were extracted from the Scopus online database).

**Figure 5.** Most productive institutions. (Bibliometric data were extracted from the Scopus online database).

For China, the most cited "Article" (60 times) was addressed on the application of near-infrared spectroscopy to agriculture and food analysis; moreover, the authors marked the information sharing mode between the central database and end-user by using network technology and concentrating valuable resources [40]. One "Review" is reported and is focused on how a transfer program on metrology for safe food and feed in developing economies was started at the International Bureau of Weights and Measures to allow national metrology institutes or designated institutes to work together to strengthen their national mycotoxin metrology infrastructure, through a description of an application of an accurate characterization of a pure aflatoxin B1 material to avoid calibration errors. It is worth noting that mycotoxins, secondary metabolites produced from microfungi in some conditions, may represent a health threat for crops, food and feed and hence for humans through the carry-over process [41].

For Italy, among 32 documents, two reviews were relevant in the approach. One opens the door to the role of incurred materials in method development and validation in order to account for food processing effects in food allergen analysis [42]. The second is a stakeholders' guidance document for consumer analytical devices with a focus on gluten and food allergens [43]. In particular, the recommendations are based on the current known technologies, analytical expertise, and standardized AOAC INTERNATIONAL allergen community guidance and best practices for the analysis of food allergens and gluten [43].

The reported institutions have produced at least six documents. The most productive Institution is the National Institute of Metrology China with 17 documents. The most cited "Article" is a study describing an approach for the identification and determination of arsenic species in kelp extract [44]. Other important studies among the reported articles are the study of Guo et al. [45] on certified reference materials and metrological traceability for mycotoxin analysis and the study of Xue et al. [46] on reference material for the quantitative detection of *Escherichia coli*. The work of Sun et al. [47] on the comparison of maximum residue levels and the standard analytical method for pesticides in tea is another relevant example. It is also worth mentioning the recent work of Joseph et al. [48], a key comparison study on organic solvent calibration solution-gravimetric preparation and the value assignment of trans-zearalenone in acetonitrile.

For the National Institute of Advanced Industrial Science and Technology, the most cited paper is the paper of Zhu and Chiba [49] on the determination of cadmium in food samples by ID-ICP-MS with solid phase extraction for eliminating spectral interferences. Two reviews are also reported: one on a proficiency test in Japan for the elements in tea-leaf powder [50], the other on the assessment of technical problems in the analysis of inorganic elements in squid through proficiency testing [51].

One hundred and five terms in total were obtained from the literature quantitative analysis on publications, and they are visualized as a term map in Figure 6. Terms such as quality control, units of measurement, calibration, measurement/s, standard/s, reference standards, certified reference materials, and reference material appear as the top-recurring keywords: this shows the integrated research in the food and agriculture area, which is based on the control quality procedure and metrology principle (Table 1).

**Figure 6.** Term map for relationships between metrology, agriculture, and food research. The number of publications was represented by bubble size. The citations per publication (CPP) were given by bubble color. Two bubbles that are closer to each other reflect that the terms coappeared more frequently. (Bibliometric data were extracted from the Scopus online database and elaborated by VOSviewer software).


**Table 1.** The top-recurring terms on the relationships between metrology, agriculture and food research. (Bibliometric data were extracted from the Scopus online database and elaborated by VOSviewer software).

Analytical methods, reference materials, reference standards, calibration, and proficiency testing represent the key elements for quality control and ensuring the accuracy of

results, as indicators of metrology traceability. The traceability of routine analysis is critical for accurate measurements.

#### **4. Conclusions**

The proposed search methodology based on a quantitative literature analysis represents a useful and potent tool to identify emerging research directions, collaboration networks, research infrastructures, and authors that are more active in the selected area of research. This may provide suggestions for more in-depth literature searches. It can be concluded that the application of metrology could provide an important contribution to the overall frontier research in agriculture and food areas worldwide, aligning investigation, research, and innovation with society's values, needs, and expectations. The proposed perspective could represent a starting point for indicating the importance of metrology in the explored areas of research, which also impact health, traceability, sustainable economy, and safety in the agro-food system, among other things.

**Author Contributions:** Conceptualization, A.D. and M.L.; methodology, A.D. and M.L.; investigation, A.D. and M.L.; writing—original draft preparation, A.D. and M.L.; writing—review and editing, A.D., E.B.S., G.L.-B., A.S. and M.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


### *Perspective* **Antioxidant Properties of Bee Products Derived from Medicinal Plants as Beekeeping Sources**

**Alessandra Durazzo 1,\*, Massimo Lucarini 1, Manuela Plutino 2, Giuseppe Pignatti 3, Ioannis K. Karabagias 4,5, Erika Martinelli 6, Eliana B. Souto 7, Antonello Santini <sup>8</sup> and Luigi Lucini <sup>6</sup>**


**Abstract:** Plant species are fundamental source of nectar in beekeeping since bees access nectar and pollen from flowers. Consequently, bee products are strongly linked to the bee foraging flora source, and, depending on this, they acquire defined features, including their health and medicinal properties. Medicinal plants contribute greatly to increase the beneficial properties of bee products, such as honey, pollen, royal jelly, and propolis. Bee products represent a potential source of natural antioxidants that can counteract the effects of oxidative stress underlying the pathogenesis of many diseases. The antioxidant properties of bee products have been widely studied and there is an abundance of information available in the literature. Notwithstanding, the uniqueness of the presented perspective is to provide an updated overview of the antioxidant properties of bee products derived from medicinal plants as beekeeping sources. This topic is divided and discussed in the text in different sections as follows: (i) beekeeping and the impacts of environmental factors; (ii) an overview of the role of medicinal plants for bee products; (iii) definition and categorization of the main medicinal bee plants and related bee products; (iv) the study approach of the antioxidant properties; (v) the conventional and innovative assays used for the measurement of the antioxidant activity; and (vi) the antioxidant properties of bee products from medicinal plants.

**Keywords:** antioxidant properties; bee products; honey; propolis; plant sources; medicinal plants

#### **1. Introduction**

Bee products represent a potential source of natural antioxidants, including phenolic acids, flavonoids, and terpenoids as well as numerous other phytochemicals, which are capable of counteracting the oxidative stress effects underlying the pathogenesis of many diseases [1]. The main action of the antioxidants is based on the capability to inhibit oxidation processes, thus reducing the production of free radicals, which result in triggering a chain reaction, which may cause harmful cellular alterations [2]. Reactive oxygen species (ROS) are produced by living organisms as a result of the normal cellular metabolism and environmental factors. The ROS are highly reactive molecules involved in many cellular signaling pathways, and can damage cell structures, such as carbohydrates, lipids, nucleic

**Citation:** Durazzo, A.; Lucarini, M.; Plutino, M.; Pignatti, G.; Karabagias, I.K.; Martinelli, E.; Souto, E.B.; Santini, A.; Lucini, L. Antioxidant Properties of Bee Products Derived from Medicinal Plants as Beekeeping Sources. *Agriculture* **2021**, *11*, 1136. https:// doi.org/10.3390/agriculture11111136

Academic Editors: Ilaria Marotti and Ângela Fernandes

Received: 22 July 2021 Accepted: 9 November 2021 Published: 13 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

acids, and proteins, and consequently alter their functions [3,4]. Oxidative stress is defined as a condition resulting when the critical balance between free radical generation and antioxidant defenses is unfavorable [5–7]. The state of oxidative stress could be related to various degenerative diseases, such as atherosclerosis, cancer, neurological disorders, diabetes, and cardiovascular disease [8,9].

The antioxidant compounds contained in bee products have different mechanisms of action causing the decrease of the adverse consequences of reactive oxygen and nitrogen species, which lead to oxidative stress. They can inhibit the enzymes responsible for producing superoxide anions and metal chelation, break the radical chain reactions, and play a preventive role inhibiting the formation of the reactive oxidants species [10]. The antioxidant properties of bee products have been widely studied for their relevant interest [11,12]. The current trend of interest in this topic is evident by the substantial amount and typology of the existing published research papers on bee products and antioxidants. For example, a search on honey throughout the Scopus online database was carried out by means of the string TITLE-ABS-KEY (honey\* AND "antioxidant property\*" OR "antioxidant capacity\*" OR "antioxidant assay\*"). The "full records and cited references" were exported and processed using the VOSviewer software (version 1.6.16, 2020; www.vosviewer.com, accessed on 6 June 2021) [13–15]. The search returned 713 publications covering the time range from 1996 to 2021, and a total of 559 terms were identified and visualized as a term map in Figure 1. Figure 1 allows for the identification of the main terms to be correlated to research on the relationship between antioxidant properties and honey, and also identifies the main existing research lines focused on this topic. It is interesting to observe that among the top-recurring keywords, compounds such as phenols, flavonoids, phenol derivative, and polyphenols appear.

**Figure 1.** Term map for the relationship of honey and its antioxidant properties research. Bubble size represents the number of publications. Bubble color represents the citations per publication (CPP). Two bubbles are closer to each other if the terms co-appeared more frequently (bibliometric data were extracted from the Scopus online database and elaborated by the VOSviewer software).

Medicinal plants contribute greatly to increase the beneficial properties of bee products e.g., honey, pollen, royal jelly, and propolis, and have the potential to produce bee products with higher bioactivity. The value of honeybee products is strictly related to the plants that attract honeybees.

The uniqueness of this perspective is to provide an updated overview of the antioxidant properties of bee products derived from medicinal plants as beekeeping sources. The topic is discussed as follows: (i) beekeeping and the impacts of environmental factors; (ii) overview of the role of medicinal plants for bee products; (iii) definition and categorization of main medicinal bee plants and related bee products; (iv) study approach of the

antioxidant properties; (v) conventional and innovative assays used for the measurement of the antioxidant activity; and (vi) antioxidant properties of bee products from medicinal plants. To our knowledge, these features have not been studied together in previous perspectives in the literature.

#### **2. Bee Products, Medicinal Plant, and Environment: An Overview**

An overview of the linkage of beekeeping, bee products, medicinal plants, and environment is delineated in the following sub-sections by underlining the key role of medicinal plants for bee products. The following topics are explored: (i) beekeeping and the impacts of environmental factors; (ii) overview of the role of medicinal plants for bee products; and (iii) definition and categorization of the main medicinal bee plants and the related bee products. The sizable contribution of medicinal plants to health properties of bee products is notable.

#### *2.1. Beekeeping and the Impacts of Environmental Factors*

Bees are considered significant pollinators due to their effectiveness and wide diffusion worldwide. Bee pollination provides excellent value to crop quality and quantity, improving global economic and dietary outcomes [16].

Honeybees (*Apis mellifera*) are social species and represent one of the most important pollinators for agricultural systems [17,18]. Humans have managed honeybees for thousands of years, and developed bee breeding for all continents, mostly the United States and Europe [19,20].

Bees have been used to produce honey and play a mandatory role in pollination since the time of Ancient Egypt populations [21,22]. Many centuries later, in the seventeenth and eighteenth centuries, the improvement of beekeeping techniques made it possible to maintain large bee colonies giving rise to modern apiculture [23].

In the United States and Europe, Bruckner et al. [24] showed that beekeepers faced significant seasonal problems, namely high mortality occurred over the winter. In the United States, beekeepers contain 30% colony losses each winter. Other studies showed that seasonal floral resources, insecticide use, and the availability of natural environmental habitat are major drivers of bee health and abundance [25].

Honeybees are constituted by over 20,000 species and each colony contains thousands of individuals [26,27]. In Central Europe, the number of managed honeybee colonies have decreased since the 1960s [28,29].

This decrease in number could be attributed to bee activities, their survival and antropic impact on the eco systems are closely connected; humans depend on bees for ecosystem services and the bees depend on the antropic activities for their survival [30,31]. On the other hand, antropic impact and environmental pollution are main issues regarding honeybee survival, and in this regard, it should be considered that the life of many plant species depends on bee pollination. The reduction of bee colonies poses a serious risk to many plant species survival, and can also be considered a biomarker for human health [32,33].

In an ecosystem global vision, recognizing several ecosystem services provided by bees is necessary, and so are the large variety of ecosystem services to humans, such as pollination, provision, regulation, and equilibrium [34]. If only one ecosystem service is considered, sharp declines in the provision of other ecosystem services may occur [35].

The honeybee is one of the most studied among all animals. This research has been almost entirely developed on the European honeybee *Apis mellifera*. According to the Food and Agriculture Organization (FAO) [36], only 11 new honeybee species have been recorded and identified in the past 15 years (Table 1).


**Table 1.** Eleven honeybees species as recorded worldwide [36–38].

#### *2.2. The Role of Medicinal Plants for Bee Products: An Overview*

Plant species are fundamental in beekeeping as a source of nectar. The species visited by bees provide the honey with particular features: pleasant taste, typical color, and pharmaceutical beneficial properties. Honey is an exclusive vegetable product. About 80% of the world's population choose plant-based extracts for basic health care. The local population in many areas of the world uses medicinal plants holistically, and considers them an important source for the prevention and treatment of diseases, especially in low income countries where a conventional pharmacological approach may be difficult [39–42]. Honey is considered a therapeutic food, which possesses pharmacological activity [43].

The flora existing in the bees' environment is important for beekeeping since the bees collect nectar and pollen from flowers. The importance of flora in beekeeping has been observed by various authors around the world [44–46]. Plant species differ from place to place, in their flowering duration, due to climate, topography and agricultural practices. Knowing the type, density, and quality of bee flora are among the crucial factors for the success and productivity of beekeeping [47].

Forests, grasslands, agrophytocenoses (e.g., orchards, vineyards, flower crops), medicinal plant plantations, and aromatic herbs are frequently visited by bees to search for melliferous plants [48].

According to Bakour et al. [49], most beekeepers prefer plants characterized by several and numerous beneficial properties for the well-being of humans as antioxidant, antiinflammatory, antifungal, antidiabetic, diuretic, having effects in the cure of different types of cancer, and also in the neurodegenerative, cardiovascular, and gastrointestinal tract diseases. The medicinal properties of bee products are also dependent on their botanical sources, as previously reported [50]; the several botanical origins provide bee products with numerous medicinal properties and products with therapeutic features for consumers.

Some of the most popular bee products for positive human health characteristics have been reported [51–53]:


The aforementioned products are strongly linked to the vegetable source from which the bees acquire their properties, including health and medicinal properties.

The importance of medicinal plants for beekeeping not only refers to the possibility of obtaining bee products from these species. In beekeeping, for some time, medicinal plant species have also been used as an alternative to pesticides [54–56]. According to Khan et al. [57], many honey bee pathogens are contrasted with medicinal plants by beekeepers. Several medicinal plants are effective against fungi, mites as *Varroa* spp., and bacteria. Nguyen et al. [58] investigated physicochemical and viscoelastic properties of honey from medicinal plants, i.e., Tulsi, Alfalfa, and two varieties of Manuka honey derived from medicinal plants.

In the next section, bee products derived from medicinal plant sources are described and explored in detail.

#### *2.3. Definition and Categorization of Main Medicinal Bee Plants and Related Bee Products*

Many plants considered mainly as food or raw material sources have some special beneficial health effect [59–64]. Sage (*Salvia* spp.) leaves, for example, are the basis for an herbal tea, which can be used for medicinal purposes, but these are also used in food preparation as spices and seasonings (as an aromatic plant), while the essential oil is used in cosmetics (e.g., soaps, and toothpaste). In this sense, the term 'medicinal' is often understood in a wide sense, and includes several overlapping uses as herbal teas, spices, food, raw material, dietary supplements, and cosmetics containing extracts of derived compounds from plants [65].

The World Checklist of Useful Plant Species contains more than 40,000 taxon names from more than 400 families and 6000 genera with a documented human use [66]. Medicinal plants, both for human and veterinary use, account for 26,662 species, and this number constantly increases with research on uses in folk or traditional medicine systems and with the addition of new plant species. The exact number of species used as medicinal material in Europe is difficult to ascertain because of the limited amount of used material (which escape from trade catalogues), the origin from several (often undetermined) plant species and lack of documentation for local uses. About 2000 taxa commercially available are sources of medicinal and aromatic plant material, and two-third of them are represented by species native to Europe [65]. In particular, Germany (1500 taxa, 600 native), France (900 taxa, 450 native), and Spain (800, 600 native) are the leading countries in the trade of medicinal and aromatic plants in Europe.

The demand for wild-collected plants species is increasing worldwide, and has become a risk for the conservation and preservation of the natural resources [67]. Using data filtering from global checklists of medicinal plants, about 400 medicinal plant species native to Europe have been included in the European red list of medicinal plants [68]. For the species in which sufficient data are available, the 2.4% were assessed as threatened and the 4.5% near threatened, following International Union for Conservation of Nature (IUCN) Red List Categories. The highest number of species was found in the Mediterranean area and in mountain areas (e.g., Alps, Pyrenees, Massif Central, Balkan Peninsula), with a similar pattern for endemic plant species.

In the same study, the major threats have been identified as wild plant collection and loss of habitat (respectively, 26% and 30% of all species), connected to human impact, the so called anthropic effect (e.g., livestock farming, recreational activities, tourism, use of chemicals, pollution, and urban development).

Bee foraging activity on plants is dedicated to the search of nectar and honeydew as carbohydrate sources, pollen (as a protein source), and resins to produce propolis (for antimicrobial and defense purposes) [69]. Pollen is collected by the bees during their visits on the plant flowers and it is stored in pellets as a protein source into the hive. The botanical origin of both honey and pollen depends on the flora surrounding the area of foraging and influences physico-chemical, functional, and sensory properties of bee products [70–72].

More than half of the medicinal plant species of the European checklist might be considered as relevant nutrition sources for bees (Table 2).



Species belonging to the *Lamiaceae* (37 medicinal plant species), *Orchidaceae* (30), *Rosaceae* (26), and *Asteraceae* (16) plant families are the main species. *Lamiaceae* plant genera (e.g., *Lavandula*, *Thymus*, *Teucrium*, *Salvia*, *Stachys*) are visited by bees mainly as nectar sources, while *Rosaceae* (e.g., *Malus*, *Prunus*, *Rubus*, *Crataegus*) and *Asteraceae* (e.g., *Aster*, *Taraxacum*, *Tussilago*, *Tanacetum*, *Helichrysum*) are for both nectar and pollen foraging, although many plant species offer both nutrition resources to bees. On the opposite, genera of the *Orchidaceae* family (e.g., *Anacamptis*, *Dactylorhiza*, *Ophris*, *Orchis*) have a limited value for foraging, although these species attract wild bees (e.g., bumble bees of the genus Bombus) with their flowers [75]. Overall, many plants of these families have great relevance

for oligolectic wild bees, which are adapted to collect pollen only from a small number of plant species [76].

By extending the analysis from the European medicinal plant checklist to some mainly cultivated or non-native medicinal plant species (Table 3), several plant taxa are of importance as sugar or pollen resources for bees.

**Table 3.** Some examples of plant taxon (family, genus) of cultivated/non-native medicinal plants in Europe, in relation to bee foraging sources (nectar, pollen). Plant data from Wichtl [77], bee preferences for pollen and nectar from MLR, [73], MAA, [74].


Some examples include tree species of the families *Rutaceae* (e.g., *Citrus*), Rosaceae (e.g., *Prunus*), *Myrtaceae* (e.g., *Eucalyptus*), shrubs of *Fagaceae* (e.g., *Corylus*), *Eleagnaceae* (e.g., *Hippophae*), and herbaceous species *Asteraceae* (e.g., *Helianthus*), *Apiaceae* (e.g., *Coriandrum*, *Levisticum*, *Pimpinella*), which are planted for different purposes (fruit crop, timber, vegetable) or belong to an anthropic habitat. An example of an interesting new medicinal and melliferous plant, the *Perilla frutescens*, an annual herb originating from China, Japan, India, Thailand and Korea, and belonging to the mint family (*Lamiaceae*) also grows in Italy, and is described and discussed by Barbieri and Ferrazzi [78].

In Figure 2, some medicinal plants growing in Italy are shown in relation to their foraging importance for bees (e.g., as sources of nectar, pollen, propolis or honeydew).

Apart from the nectar, bees also collect honeydew as a sugar source, if available in their foraging area. The main sources of honeydew are forests and conifer trees, which originate in the secretions from the living part of the plant (e.g., the leaves) or from sap-sucking insects [71]. In Europe, honeydew honey originates mainly from fir (*Abies alba*), spruce (*Picea abies*), and *Pinus* (e.g., *Pinus halepensis*, *P. brutia*) trees, but also from *Salicaceae* (*Salix*, *Populus*), *Fagaceae* (*Castanea*, *Quercus*), *Oleaceae* (*Fraxinus*, *Olea*), *Tiliaceae* (*Tilia*), *Betulaceae* (*Betula*) and *Sapindaceae* (*Aesculus*) [71]. Due to the variety of nectar and honeydew sources in natural or artificial habitats, a wide range of different types of honey can originate [79].

Various floral honeys are regarded as medicinal honeys with high polyphenol contents. Manuka dark-colored honey, for example, originates from *Leptospermum scoparium* and *L. polygalifolium*, shrubs native to Australia and New Zealand of the *Myrtaceae* family [80,81]. Another example are the honeys from *Acacia ehrenbergina* (*Fabaceae*) and *Ziziphus* spinachristi (*Rhamnaceae*), trees native to some areas of Africa and Asia, which show high phenolic contents [82]. Propolis originates from collected vegetable material by bees and is mixed with wax. Main sources for propolis production by bees are restricted to a small number of species, which are typical for specific geographic areas [79,83]. In Europe and North America, tree species of the genera *Populus* (e.g., *Populus tremula*, *P. nigra* in Europe, *P. deltoides* and *P. trichocarpa* in America) and *Betula*, are known as resin resources for

bees, while in tropical and subtropical areas, *Dalbergia* and *Acacia* (*Fabaceae*), *Macaranga* (*Euphorbiaceae*), *Mangifera* and *Rhus* (*Anacardiaceae*), and *Baccharis* (*Asteraceae*), are the main sources. In the Mediterranean areas, where *Populus* species might be less frequent, the source of resins for propolis are the *Cupressus* sempervirens and the *Juniperus phoenicea* [84].

**Figure 2.** Examples of medicinal plant sources for bees: (**A**) *Robinia pseudoacacia* (nectar), (**B**) *Tilia cordata* (nectar, pollen, honeydew), (**C**) *Rubus ulmifolius* (nectar), (**D**) *Eucalyptus camaldulensis* (nectar, pollen, honeydew), (**E**) *Populus nigra* (propolis), (**F**) *Pinus pinaster* shoot heavily infested by *Toumeyella parvicornis* (pine tortoise scale, honeydew).

#### **3. Bee Products from Medicinal Plants: Antioxidant Properties Measurements**

Natural products, including bee products, which often contain medicinal plants containing compounds, such as honeydew secretions of *Abies*, *Betula*, *Castanea*, *Fraxinus*, *Pinus*, *Quercus*, *Rosemary*, *Thymus*, *Tilia*, and other species, are particularly appreciated by consumers for therapeutic uses as an alternative to drugs [85]. Antioxidants are sourced from the plant species. Medicinal plants generally recognized as having potential beneficial value can therefore be utilized to obtain honey with greater bioactivity and bioavailability. Nowadays, apitherapy has gained much attention from both consumers and researchers. In principle, apitherapy is a theory of alternative medicine that utilizes bee products, such as honey, pollen royal jelly, propolis, and bee venom, for medicinal purposes. However, it remains scarce and not exactly known whether treatments with bee products are safe and how the possible health risks of using such products can be minimized [85]. Proponents of apitherapy make claims for its health benefits, which in contrast, are unsupported by traditional medicine [50,86,87].

In this context, the antioxidant properties of bee products can be considered as an expression of the *melliferous* medicinal plants' therapeutic potential. Updates and considerations on the approach towards antioxidant properties have been mentioned; nonetheless, the conventional and innovative assays for the assessment of antioxidant properties and the remarkable antioxidant properties of bee products from medicinal plant cannot be neglected.

#### *3.1. Study Approach of the Antioxidant Properties: Updates and Considerations*

The combined action of bioactive compounds, nutrients, and nutraceuticals represents the first step to study antioxidant properties and can be regarded as an indicator of the "health properties" of the food matrices [88,89].

The diversity of the chemical structure of compounds, their possible interactions, and their different mechanisms of action and biological role, make difficult the assessment of a single, adequate, and reliable procedure for the determination of antioxidant properties. Antioxidant properties are an expression of the interactions between bioactive molecules and other components in terms of both the potential health benefits of food and can be viewed as a screening method for interpretation and supporting further research. Extraction, antioxidant capacity measurements, and expression of the results can be viewed as the three key steps in the evaluation of the antioxidant properties. A study by Durazzo et al. [90] reported as the main workflow in research approach of the antioxidant properties three main steps, namely: (i) the development of a system as model study of the compounds' interactions; (ii) the investigation of extractable and non-extractable compounds; (iii) the behavior study of bioactive compounds-rich extracts.

Nowadays, the distinction between extractable and non-extractable antioxidants has been recognized as a fundamental aspect in the definition of the healthy properties in terms of the prevention of diseases [91,92]. In particular, the distinction between extractable and non-extractable antioxidants has achieved a shared consensus in the scientific community.

Indeed, new research directions point to the exploitation of new and unconventional sources for antioxidants and to the identification of new possible applications.

This research on the antioxidant properties should be based on an integrated and multidisciplinary approach, resulting from a combination of studies in several areas, such as nutrition, food chemistry, phytochemistry, and medicine. Innovative design of study research includes green procedures and sustainable technologies, and the joined up use of statistical methods, such as chemometrics.

An overall challenge is the development of dedicated databases for the antioxidant properties and the inclusion into harmonized databases; these are studies currently being carried out.

The inclusion of extractable and non-extractable compounds in current comprehensive and harmonized databases have been developed in the eBASIS BioActive Substances in Food Information System [93–95]. The development of search protocols and data collection systems have allowed to obtain new quality evaluated data on extractable and non-extractable antioxidants, used for the expansion eBASIS, leading thus, to a valuable unique data resource [96]. A total of 437 datapoints on the composition of extractable and/or non-extractable compounds were added into the database. This update of eBASIS can be viewed as the first examples of building a database dedicated to antioxidant properties. This eBASIS 'expansion provides a new and unique tool for dietitians, nutritionists, and researchers for a great range of uses, e.g., dietary assessment, epidemiological studies, and exposure studies [96].

In this context, the study of Pellegrini et al. [97], by summarizing 25 years of investigations on antioxidants research in foods and biological fluids, remarked how the availability of well-constructed Total Antioxidant Capacity databases deserves attention and must be considered. Moreover, the same authors highlighted how the appropriate use of Total Antioxidant Capacity measurement both in food and in vivo can still support interpretation of complex phenomena and can be viewed as a useful tool, for instance, for the sample screening when making a quick decision toward in-depth research investigations [97].

#### *3.2. Antioxidant Properties Assessment: An Overview of Conventional and Innovative Assays*

A variety of assays aimed at evaluating the dietary antioxidant properties have been proposed, although a reliable and commonly accepted assay has not been so far identified [98]. Overall, the methods available can be grouped into three major classes, namely in vitro, cellular, and in vivo assays. In vitro chemical assays are the most frequently used, because these are cheap and have high-throughput, but their prediction ability has been questioned in recent years [99,100]. Cell-based assays are considered a middle ground between the in vitro and in vivo tests (the latter posing ethical issues, high cost, and limited throughput) [101]. However, cellular antioxidant assays still suffer from poor standardization (e.g., differences in cell line, radical generators, fluorescent probes, etc.), making the reported results difficult to be compared across the available studies [102]. Moreover, some authors claimed that cell culturing itself may induce oxidative stress as a consequence of culture conditions, hence inducing cell acclimation and, thus, overestimating the antioxidants efficacy [103].

In this complex framework, in vitro tests still represent the most frequently used antioxidant assay methods in food science. Among others, they include the DPPH (α,αdiphenyl-β-picrylhydrazyl), the ABTS (2,2 -azinobis(3-ethylbenzothiazolin-6-sulphonate), FRAP (ferric reducing antioxidant power), the CUPRAC (cupric-ion reducing antioxidant capacity), and ORAC (oxygen radical absorbance capacity) assays, [98,104–106]. The measurement of the ability of a food or a food component to scavenge specific free radicals or to reduce a target molecule are the base for all the above mentioned methods. Differences in their principles, mechanisms, experimental conditions, and in how their end points are measured occur, and for this reason the use of several methods to estimate and/or determine the antioxidant properties is suggested. Procedures and applications for these assays should be considered for standardization [107]. Moreover, one of the main concerns on the above mentioned assays is that these are not carried out under physiological conditions, and thus, their ability to predict in vivo effects has been questioned. Notwithstanding, most of the scientific literature on bee product antioxidant capacity is referred to in vitro tests, as with other foods, with the exception of a recent paper that investigated the cellular antioxidant capacity of different Moroccan Zantaz honey samples [108].

In general, in vitro assays account for different antioxidant mechanisms that include hydrogen atom transfer (HAT), and single electron transfer (SET) rather than chelation of transition metal ions. ORAC is probably the most representative HAT-based assay, while DPPH, CUPRAC, ABTS and FRAP are SET-based methods. However, the ability to chelate Cu2+ and Fe2+, a key initiation step in the oxidation processes, has also been considered in the context of the foods antioxidant properties [109].

Regarding the antioxidant properties of bee products, as with any other food, the limitations in these assays must be considered. Moreover, it must also be considered that in many situations, both HAT and SET occur simultaneously in vivo, and that antioxidant compounds may also act indirectly, via the regulation of antioxidant enzymes.

#### *3.3. Antioxidant Properties of Bee Products Relates to Foraging on Medicinal Plants*

The link between bee foraging on medicinal plants and antioxidant properties of bee products has received increasing attention. Nicewicz et al. [110] compared the antioxidant capacity of honey from urban areas vs. rural apiary, reporting that all antioxidant parameters were significantly higher in honey from rural than in urban areas. Such differences

were not ascribed to the effect of the floral composition of honey, being rather due to the location of the honeybee colonies. In recent years, while wild pollinators are declining in many landscapes, urban areas provide high plant diversity and foraging sources for bees. Therefore, a growing interest for beekeeping in cities is observed, but the quality and safety of honey produced in urban areas impacts on the consumers' concerns. Therefore, further research in this direction is needed [75].

The following sub-sections discuss the relationship between floral diversity and bioactivity of bee products, with a specific focus on each product.

#### 3.3.1. Honey

Honey (or honeybee honey) is a sugary foodstuff prepared by honeybees. Bees produce honey from the sugary secretions of plants (floral nectar) or from secretions of other insects of the *Aphids* family (honeydew) through their enzymatic activities and other biochemical processes, such as regurgitation and water evaporation. Honeybees store honey in wax structures called honeycombs, whereas stingless bees store honey in pots made of wax and resin [111]. Many centuries ago, in different civilizations, honey was used by ancient Greeks and Egyptians, and in Indian and Chinese traditional medicine both orally and topically to treat various illnesses. Traditional medicine reports uses towards stomach disorders, ulcers, skin wounds, and skin burns [112]. Several honey products have gained medical status and have been approved by the United States Food and Drug Administration (FDA) for their use in the treatment of wounds and burns [113]. It has been reported that honey has antioxidant, antibacterial, and antibiotic properties. The existence of numerous phytochemicals in different types of honey originating from plants known for their medicinal properties, such as *Thymus*, *Abies*, *Pinus*, *Castanea*, and *Rosemary* botanical species supports its antioxidant activity [114–118].

In more detail, Gheldof et al. [118] reported that buckwheat (*Fagopyrum esculentum*) honey increased human's serum antioxidant activity. Anand et al. [119] characterized the physico-chemical and antioxidant properties of *Agastache* honey produced from *Agastache rugosa* in comparison with commercial honeys sold in the Australia market. Their results confirmed that *Agastache* honey had a superior antioxidant capacity [119]. More recently, Adgaba et al. [120] studied, among others, the antioxidant and anti-microbial properties of some Ethiopian monofloral honeys, reporting average total antioxidant values of 320.3 ± 15.1 μM Fe(II)/100 g with a range of 225.4–465.7 μM Fe(II)/100 g. The same study reported relatively higher values (421.5 ± 23.4 and 465.7 ± 21.8 μM Fe(II)/100 g) for *Croton macrostachyus* and *Vernonia amygalina* honeys, respectively. Nonetheless, in a review article focusing, among others, on the antioxidant properties of monofloral honeys, several honey types produced in different countries, such as acacia (*Acacia* sp.), astragalus (*Astragalous microchephalus* Willd), linden (*Tilia* sp.), willow (*Salix* sp.), and others, were reported to provide antioxidant capacity using multiple antioxidant assays, such as ABTS, DPPH, FRAP, ORAC, and TEAC (Trolox equivalent antioxidant capacity) [121].

The antioxidant activity of the studied honeys was affected by the botanical source and the geographical origin [121].

#### 3.3.2. Bee Pollen and Its Derivatives

Bee pollen is among the honeybee products that contain nourishing nutrients, which can provide energy to humans. The health-enhancing value of bee pollen is owed to plant secondary metabolites, such as tocopherols, niacin, thiamine, biotin and folic acid, polyphenols, carotenoids, phytosterols, enzymes and other co-enzymes [85,122]. However, the studies highlighting the antioxidant, anti-inflammatory, anticariogenic antibacterial, antifungicidal, hepatoprotective, anti-atherosclerotic, immune enhancing properties need to be more extensive, concerning mainly the application of cohort clinical trials. The basic hurdle in the use of bee pollen as functional component is probably related to the broad species/specific variation in its composition [122]. Such variations may differently contribute to the properties of bee-pollen and biological activity, and thus, may affect

its therapeutic effects (positively or negatively). Notwithstanding, bee pollen has been recommended as a valuable dietary supplement [85]. Pollen antioxidant activity has been related to a wide range of botanical species, such as *Papaver rhoes*, *Chamomila recutita*, *Sinapis arvensis*, *Cistus* sp., *Trifolium* sp., *Dorycnium* sp., *Cichorium* sp., *Convolvulus* sp., *Circium* sp., *Malva sylvestris*, *Fumana* sp., *Eucalyptus camaldulensis*, *Anemone* sp., *Ononis* sp., *Asphodelus* sp., and *Quercus ilex* [122].

Moving toward bee pollen-derived products, bee bread has a similar composition to bee pollen, but with marked quantitative differences mainly related to the fermentation process, which it undergoes. For instance, bee bread delivers higher amino acids, sugar, lactic acid, and vitamin content compared to bee pollen [123,124]. In a recent study, bee bread from different regions in Greece, containing *Castanea sativa*, *Cistus* sp., *Hedera helix*, *Borago* sp., and other pollen grains belonging to the *Brassicaceae* family, showed both antibacterial and antioxidant activity [125].

#### 3.3.3. Propolis

Propolis, commonly known as the "bee glue" is a resinous mixture that honeybees produce by mixing their saliva, which contains enzymes and beeswax, with exudate gathered from different plant materials such as leaf and flower buds, stems, and bark cracks of numerous tree species. The word propolis originates from the two Greek words "pro" and "polis", which mean "defense" and "city" or "community," respectively [50]. Propolis is typically composed of 50–60% of resins and balms (including phenolic compounds), 30–40% of wax and fatty acids, 5–10% of essential oils, 5% of pollen, and approximately 5% of other components, including amino acids, micronutrients, and vitamins (thiamin, riboflavin, pyridoxine, vitamins C, and E). More than 300 compounds belonging to polyphenols, terpenoids, steroids, sugars, amino acids, and others have been identified in propolis [126].

The antioxidant activity of propolis has been determined by the use of in vitro methods, such as DPPH, ABTS+, FRAP, and ORAC [50]. Interestingly, the antioxidant activity of propolis extracts was comparable to the synthetic antioxidant butylated hydroxytoluene (BHT) or to ascorbic acid. Moreover, studies regarding the antioxidant properties of propolis, have also been carried out on cell cultures and animals. Clinical trials investigating the antioxidant effect of propolis reported a positive modulation of cardiovascular disease markers, a mitigation of chemotherapy side effects, as well as neuroprotective effects [127–129].

Similar to other bee products, plant sources are related to the profile of bioactive compounds and antioxidant properties of propolis [130,131]. More specifically, in regions with a large diversity of trees, bees may also gather resin from flowers in the genera *Clusia* (*Clusia* L.) and *Dalechampia* (*Dalechampia* L.), which are the major plant genera that produce floral resins to attract pollinators [132]. Clusia resin contains polyprenylated benzophenones [133,134]. In some areas of Chile and Brazil, propolis contains viscidone, a terpene from *Baccharis* (*Baccharis* L.) shrubs, and prenylated acids, such as 4-hydroxy-3,5 diprenyl cinnamic acid [135,136].

#### 3.3.4. Royal Jelly

Royal jelly is a mixture of secretions from the mandibular and hypopharyngeal glands of bees of the *Apis mellifera* species, representing the major food source for the queen honeybee [50]. Concerning its composition, royal jelly is an emulsion of proteins, sugars, and lipids in water. Moreover, it contains approximately 1.5% (*w/w*) of minerals (mainly copper, zinc, iron, calcium, manganese, potassium, and sodium) and considerable amounts of flavonoids, polyphenols, and vitamins (biotin, folic acid, inositol, niacin, pantothenic acid, riboflavin, thiamine, and vitamin E). Among the flavonoids, the flavanones (hesperetin, isosakuranetin, and naringenin), flavones (acacetin, apigenin, and its glucoside, chrysin, and luteolin glucoside), flavonols (isorhamnetin and kaempferol glucosides), and isoflavonoids (coumestrol, formononetin, and genistein) are the most abundant [137].

The antioxidant activity of royal jelly, in terms of DPPH, hydroxyl and superoxide radical scavenging, has been reported in the literature [138,139]. In this regard, both in vitro and in vivo tests have been conducted, whereas less information has been reported from clinical trials. In a recent study carried out by Pourmoradian et al. [140], the positive impact of royal jelly consumption on the parameters associated with diabetes and oxidative stress in people affected by Type 2 diabetes mellitus has been postulated. On the other hand, studies on rats and rabbits reported that royal jelly intake can be associated with antioxidant and neuroprotective effects [141,142]. This is consistent with the antioxidant activity of monophosphate nucleotides and peptides isolated from royal jelly [143].

More generally, the antioxidant activity of royal jelly may be differentiated compared to honey, bee pollen and propolis. However, the specific contribution of the botanical species available for foraging on the actual functional properties of royal jelly is still poor.

#### **4. Conclusions and Future Directions**

It is clear that medicinal plants can contribute to the antioxidant activity of bee products along with the honeybee contribution as a living organism; the antioxidant properties can be regarded as an indicator of the *melliferous* medicinal plant's potential. In this context, more research should be focused on bee products obtained from the broad range of medicinal plants and on the identification of the possible relationships between the bioactive components, which are present in plant parts and their nectars as well as the bee products.

At the same time, research and clinical trials should be conducted on humans to assess the relationship between the consumption of bee products and the aiding or treatment in health disorders. In this way, the potential use of bee products in phytomedicine (as an alternative to drugs) could be better substantiated by the scientific evidence. The complementary use of the nanotechnologies [144,145] opens new directions and new frontiers. For instance, Neupane et al. [146], by developing Himalayan honey-loaded iron oxide nanoparticles, showed that the biological activity of Himalayan honey was enhanced significantly after loading into iron oxide nanoparticles. Sarhan and Azzazy [147] developed biocompatible, antimicrobial crosslinked honey/polyvinyl alcohol/chitosan nanofibers, which hold potential as an effective wound dressing source. These aspects are relevant and trigger an additional interest for research to obtain a greater bioavailability and efficacy of bee products in the field of health, including anti-COVID-19 possible beneficial effects [148–151], also increasing the interest in studies that are carried out to assess the safety aspects of nanoformulations, which are indeed new frontiers to explore.

**Author Contributions:** All authors have made a substantial contribution to the writing and revision of work, and approved it for publication. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


MDPI St. Alban-Anlage 66 4052 Basel Switzerland Tel. +41 61 683 77 34 Fax +41 61 302 89 18 www.mdpi.com

*Agriculture* Editorial Office E-mail: agriculture@mdpi.com www.mdpi.com/journal/agriculture

MDPI St. Alban-Anlage 66 4052 Basel Switzerland

Tel: +41 61 683 77 34 Fax: +41 61 302 89 18

www.mdpi.com