*2.3. Biological Activities and SAR Investigation*

Compounds **1**–**3**, isolated from the bacteria *Pantoea* sp., were assayed for their antimicrobial and insecticidal activities against human pathogenic microorganisms and *Ae. aegypti* larvae (Table 2). The bioassays showed that compound **1** exhibited minimal larvicidal and antibacterial activities. In contrast, compound **3** demonstrated significant larvicidal activity with a mortality rate of 67.3% and antibacterial activity with an MIC of 16 μg/mL against methicillin-resistant *S. aureus*. Based on the observed differences of these compounds in the bioassays, we evaluated analogues of **1***–***3** to gain more insight into the structure-activity relationship (SAR) of this class of compounds.


**Table 2.** Antimicrobial and insecticide activities of compounds **1**–**3** and synthetic analogues **4**–**39.**

*<sup>a</sup>* Postitive control: Oxacillin for *S. aureus*; fluconazole for *C. albicans* and *T. rubrum*; vancomycin for MRSA and rotenone at 1 μg/mL for larvicidal activity. −: MIC >256μg/mL, considered not active. nd: Not determined. \* natural product.

The synthetic compounds were evaluated for their larvicidal and antimicrobial activities (Table 2). First, the results showed that the best antimicrobial activities were obtained using compounds **1**, **3**, **7**, **19**, **22**, **23**, **29**, **30**, **34**, and **35** (MIC ≤64 μg/mL against MRSA and/or *T. rubrum*) with the highest antimicrobial activity being observed with compounds **30** (MIC <8 μg/mL against MRSA and *T. rubrum*) and **7** (MIC = 8 μg/mL against MRSA). Compounds **19** and **23** only showed activity against *T. rubrum*, whereas compounds **1**, **3**, and **7** only showed antibacterial activity against MRSA. Finally, compound **29** was the only compound to demonstrate MICs ≤64 μg/mL on all tested pathogenic strains. Interestingly, compounds **1**, **3**, **7**, **23**, and **35** showed antibacterial activity against MRSA only and not *S. aureus*. Most of these compounds contain a carboxyl group and a saturated alkyl chain with 12 to 14 carbons. Similar biological activities were observed irrespective of the stereochemical

configuration of the free acids (**22**–**35**). On the other hand, given the antibacterial activities of compounds **8** and **16** against MRSA, the D configuration appears deleterious for lipoaminoesters.

Regarding larvicide potential, compounds **1**, **3**, *ent***-3**, **19**, **23**, **24**, **27**, **31**, and **32** were the most active with mortality rates ranging from 26.5% to 71.1% at a concentration of 10 ppm. These compounds all possess long-chain acyl groups (mostly C16 and C18). Moreover, the compounds prepared from D-phenylalanine (*ent***-3**, **27**, **31**, and **32**) were more active than their L-phenylalanine counterparts (**3**, **20**, **24**, and **25**). We next evaluated the effect of alkyl chain unsaturation on larvicide potential and found that unsaturated molecules were more active than saturated ones. For example, the mortality rate of compound **27** was 23.3%, while it was 71.1% for *ent***-3**. Likewise, the mortality rate of compound **33** was 2.3% while it was 53.3% for **31**. Finally, unsaturated compounds with the *Z-*configuration were found to be more active than those with the *E-*configuration (i.e., compounds **24** and **31** compared to **25** and **32**).

We also investigated the effect of amino acid replacement on compound bioactivity. In general, the methyl esters of tyrosine alanine derivatives (**18** and **19**, respectively) were not antimicrobial. In contrast, the corresponding acids (**34** and **35**) appeared to be more active than the phenylalanine analogue **3**. The ester compounds were less soluble than the acids, which might explain their weaker activity. Regarding larvicidal activity, only the methyl ester of the acylated tyrosine derivative (**19**) was active (mortality rate of approximately 30%).

In summary, compound **3** was the only compound that demonstrated larvicidal and slight antibacterial activity with mortality rates of 42.9% and 67.3% and MIC values of 16 and 64 μg/mL against *Ae. aegypti* larvae and MRSA, respectively.

#### *2.4. Molecular Networking*

A molecular network (MN) was generated to organize the tandem MS data acquired from the entire collection of entomopathogenic strains that were biologically tested in this work, including *Pantoea* sp. SNB-VECD14B. After applying appropriate data treatment to align the common features among the collection, the resulting spectral data were submitted to the Global Natural Products Social molecular networking (GNPS) platform for molecular network generation [10–12]. The resulting molecular network grouped 20,859 spectra into 2619 clusters (Figure 3). Taxonomical mapping was applied by attributing a given color code to each strain in the collection, allowing us to visually navigate through the MN and check for the presence and distribution of specific compound classes within the full collection. A cluster was identified that was primarily related to the *Pantoea* genus. Together with the results from biological screening of the 53 strains collection, we identified a series of structurally related compounds that could be responsible for the observed larvicidal activity. This was confirmed by injecting the isolated compounds **1**–**3**, which were effectively found to be present in this particular cluster. Interestingly, this cluster of lipoamino acid-related structures indicated the following: (i) There is high structural diversity in the compounds generated by the *Pantoea* genus, and (ii) some of the members of this structural family were also found to be present in other entomopathogenic strains of the collection (Figure 3).

To study in more depth the chemistry of *Pantoea* lipoamino acids, a molecular network was generated organizing the tandem MS data acquired from the *Pantoea* sp. extracts together with the MS data generated from the isolated and synthetic lipoamino acid derivatives (Figure 4). In addition to the three isolated lipoamino acids (**1**–**3**), it was possible to align six of the synthetic lipoamino acids (**4**, **8**, **9**, **14**, **20** and **22**) with features present in the crude *Pantoea* sp. extracts, thus indicating that these compounds were also naturally present in the bacterium extract (represented as squares in Figure 4). Two of the synthesized derivatives (**5** and **10**) were also found to be present in the cluster but were not detected in the natural extracts (arrow-like node in Figure 4).

**Figure 3.** (**A**) Generation of a massive molecular network from the UHPLC-HRMS<sup>2</sup> analysis of 53 entomopathogenic microorganism extracts. (**B**) Taxonomical mapping showing a cluster containing ions mainly found in the *Pantoea* sp. extract. (**C**) Cluster of lipoaminoacids with compounds **1**, **2**, and **3** corresponding to compounds isolated from the *Pantoea* sp. extract.

**Figure 4.** Cluster corresponding to natural lipoamino acids in the ethyl acetate extract of the *Pantoea* species together with synthetic lipoamino acid derivatives. The size of the nodes is proportional to the peak height of each feature within the *Pantoea* sp. extract.

The co-injection of synthetic analogues indicated that the studied extract contained additional lipoamino acid derivatives and allowed us to propagate annotations to neighboring nodes. To further annotate such compounds in the cluster, the mass spectral signature of each individual node was manually curated (Figure 5). For each feature, the molecular formula was established, annotations were made regarding adducts or complexes, and each fragmentation spectrum was treated using Sirius GUI software [17]. The results are summarized in Table S4 and allowed us to putatively annotate 16 additional compounds. Inspection of the fragmentation pattern revealed that five diagnostic fragments corresponding to the amino acid portion of the compounds could be observed within this cluster. The phenylalanine (Phe) and Phe-methyl ester [M+H]<sup>+</sup> moieties gave a typical fragment at *m*/*z* 166.09 and 188.10, respectively. Fragments at *m*/*z* 132 and 146 indicated potential leucine/isoleucine (Leu/Ile) and Leu/Ile-methyl ester [M+H]<sup>+</sup> moieties, respectively. Finally, a fragment at *m*/*z* 118 indicated a potential valine (Val) [M+H]+ moiety. It is to be noted that an additional lipoamino acid derivative bearing a tyrosine-methyl ester could also be identified in another cluster of the molecular network by co-injection with a synthetic analogue (**19**). The presence of double bonds was inferred from the calculated round double bond equivalent (RDBE). The position of the OH groups was inferred from the isolated compounds. Together with the molecular formula determination, these signature indications allowed us to propose putative partial structures for the other nodes of the cluster and highlights the significant diversity of the lipoamino acids produced by this *Pantoea* sp. strain (Figure 5).

**Figure 5.** Cluster corresponding to hypothetic analogues of lipoamino acids putatively annotated through manual curation of their MSMS spectra.
