Influence of Environmental Factors on Occurrence of Cyanobacteria and Abundance of Saxitoxin-Producing Cyanobacteria in a Subtropical Drinking Water Reservoir in Brazil

Round 1

Reviewer 1 Report

Moraes and colleagues present a well written field study on how environmental conditions influence saxitoxin production within a Brazilian drinking water reservoir. While the CCA bioplot presented in Figure 7 provides interesting information, support for quantifying STX-producing cyanobacteria using the qPCR protocol described here is less compelling.  While the primers were screened against a subset of STX-producers and -nonproducers, it is unclear how specific the primers are. Thus this manuscript could be strengthened by more focus on the cyanobacteria and environmental conditions affecting toxin production or further studies on the stx primers/qPCR protocol.

Are there stx genes that are not detected by these primers? 

Figure S1 does indicate a single amplicon is produced using the primers.  Has this amplicon been cloned and sequenced to verify its identity and specificity of the primers? 

Another issue is the presence of the stx genes versus presence of the toxin as indicated in Figure 5 and lines 340-344 and 477-485.  It would seem that an ELISA test would be more useful for water monitoring.

Lines 593-605: Based on the data provided, the $1 difference in sample analysis is not convincing enough to switch from Utermöhl technique to qPCR. 

Author Response

Reviewer #1:

Moraes and colleagues present a well written field study on how environmental conditions influence saxitoxin production within a Brazilian drinking water reservoir. While the CCA biplot presented in Figure 7 provides interesting information, support for quantifying STX-producing cyanobacteria using the qPCR protocol described here is less compelling.  While the primers were screened against a subset of STX-producers and -nonproducers, it is unclear how specific the primers are. Thus, this manuscript could be strengthened by more focus on the cyanobacteria and environmental conditions affecting toxin production or further studies on the stx primers/qPCR protocol.

 

Authors’ reply: We appreciate this comment and understand the problem: should focus be on environmental conditions or rather on the development of the sxtA assay? We think we have managed to meet both areas by presenting environmental control of cyanobacteria and their toxins, and also discussing application of the molecular assay. Reviewer #2 requested moving of the environmental variable to Supplementary Material, suggesting that these data are of less impact to the study. As mentioned, we think that we in the discussion have tried to cover both environment relations and specificity and application of the PCR assay.

 

1) Are there stx genes that are not detected by these primers?

Authors’ reply: The saxitoxin gene cluster consists of up to 33 genes and has been described for some cyanobacteria (Lyngbya wollei, Aphanizomenon sp, Dolichospermum circinale AWQC131C, Raphidiopsis raciborskii T3 and Raphidiopsis brookii D9). The primers designed in this study are specific to target only the sxtA gene which is responsible for the initiation of STX biosynthesis and appears to be carried by all saxitoxin-producing cyanobacteria.

The reason for designing a new primer set is now explained in Material and Methods with this text:

“For quantification of SXT-producing cyanobacteria, the qPCR approach by Al-Tebrineh et al. [15] targeting the sxtA gene in Anabaena circinalis and other cyanobacteria was chosen. Unfortunately, repetitive tests showed that multiple amplicons were produced by the primer set, as shown in Fig. S1. Therefore, a new primer set for qPCR was designed.”

 

2) Figure S1 [now Figure S2] does indicate a single amplicon is produced using the primers.  Has this amplicon been cloned and sequenced to verify its identity and specificity of the primers? 

Authors’ reply: To test the specificity of the designed primers, 4 saxitoxin-producing strains and 5 non-saxitoxin producing strains were analyzed with the PCR assay. The single amplicons produced (153 bp size; Figure S2) were purified and sequenced (Eurofins Genomics Company, Germany) to confirm the specificity of PCR amplification and designed primers targeting the sxtA gene. Sequence identity was determined by a Basic Local Alignment Search Tool (BLAST) of the National Center for Biotechnology Information (NCBI) database. The percent identity of the sequenced amplicons varied from 96 to 100% to the sxtA gene sequences stored in NCBI database.

 

3) Another issue is the presence of the stx genes versus presence of the toxin as indicated in Figure 5 and lines 340-344 and 477-485.  It would seem that an ELISA test would be more useful for water monitoring.

Authors’ reply: The reviewer is right that the ELISA analysis is a useful tool for monitoring of concentrations of SXT, but the ELISA method (and similar chemical approaches) cannot be used to predict if the toxins soon might be present in the water. The advantage of the qPCR assay is that it reports if there is presence of SXT-producing cells, meaning that there might soon be risk of toxins in the water, if the environmental conditions favor their growth. Thus, while ELISA is a monitoring tool, qPCR can serve as a valuable early-warning tool. It should be mentioned that none of the two methods are simple and fast field methods, since they both require operations that can realistically only be conducted in the laboratory.

 

4) Lines 593-605: Based on the data provided, the $1 difference in sample analysis is not convincing enough to switch from Utermöhl technique to qPCR.

Authors’ reply: As cited in the manuscript, the switch from Utermöhl technique to qPCR becomes economically viable only when considering the large-scale water monitoring because the qPCR method has a greater sample processing capacity (288 samples per month) and less time for obtaining the results. These characteristics make the qPCR method an excellent alternative to the traditional Utermöhl cell counting method which has less sample processing capacity (64 samples per month) and high time for obtaining the results.

Reviewer 2 Report

This is a well-written manuscript about an intriguing work that is well done. It presents new information for understanding with which environmental parameters the production of saxitoxins and cyanotoxins, in general, may be correlated. It introduces a state-of-the-art method to detect an exemplar cyanotoxin very sensitively. Therefore, it should be published soon.

The data presented are well supported by statistics; this is one of the strong sides of the work. The manuscript comprises two stories, in fact. The one is about the cyanotoxin, Saxitoxin, STX, and the stxA gene. The data presented are sound, and it was shown impressively by qPCR how sensitive the method is, i.e., even more sensitive than other methods (cell counting, pigment analyses). It is of great interest to learn that the PCR method is more sensitive, more reliable, and even cheaper than the traditional Utermöhl cell counting method. However, it still needs to be explained in as much the approach, i.e., qPCR targeting the sxtA gene, is novel. Is this work the first time ever where the approach has been used (unfortunately, there are citations 14, 15, 67-69 about similar [?] work, lines 475, 476), or is it just for the first time for a water body in Brazil (lines 472, 473)?

The other story is about the particular features found in the Brasilian water reservoir (Itupararanga reservoir). They appear to be not of general interest. They may weaken the first story about the saxitoxin because it is mostly more descriptive and likely too detailed to understand the prevailing situation where the molecular approach has been tested. The data mainly concerning measurements of local variables from different sites of the reservoir should be presented only in second place and in a much-reduced way compared to present, i.e., in a Supplementary file or separate publication dealing with local/regional data from water bodies. Therefore, the section "Results 3.1 Environmental variables" (lines 221 – 254) and Table 1 should be better moved to Supplementary information. The Results section should start with Phytoplankton composition. Figure 2 is concise and well done; it provides an impressive overview of the observed phytoplankton community and abundance of cyanobacteria. Why is it useful to compare biomass with biovolume? My non-expert impression is that biovolume is more affected by errors/wrong estimates than biomass? In the Discussion section, 4.1 Environmental variables and cyanobacteria should be moved move to 2^nd place. The Discussion should start with the more general information discussed in lines 570-574, followed by 4.2. qPCR assay, cyanotoxin, and cyanobacteria.

 

There is a minor issue with taxonomy. In line 260, Chlorophyceae is not correct. Green algae, in general, should replace it. The green algae comprise the phyla Chlorophyta (with the classes Chlorophyceae and Trebouxiophyceae) and Streptophyta (with the class Zygnematophyceae). Also, the usage of "Chlamydophyceae" is not correct; the authors probably mean the order Chlamydomonadales, which is within the class Chlorophyceae. Similarly, replace Bacillariophyceae with the more general term "diatoms" (Diatomea); the species recorded from the reservoir may, in fact, belong to several classes and not only the Bacillariophyceae.

The English style seems not to need corrections.

Author Response

Reviewer #2:

This is a well-written manuscript about an intriguing work that is well done. It presents new information for understanding with which environmental parameters the production of saxitoxins and cyanotoxins, in general, may be correlated. It introduces a state-of-the-art method to detect an exemplar cyanotoxin very sensitively. Therefore, it should be published soon.

1) The data presented are well supported by statistics; this is one of the strong sides of the work. The manuscript comprises two stories, in fact. The one is about the cyanotoxin, Saxitoxin, STX, and the sxtA gene. The data presented are sound, and it was shown impressively by qPCR how sensitive the method is, i.e., even more sensitive than other methods (cell counting, pigment analyses). It is of great interest to learn that the PCR method is more sensitive, more reliable, and even cheaper than the traditional Utermöhl cell counting method. However, it still needs to be explained in as much the approach, i.e., qPCR targeting the sxtA gene, is novel. Is this work the first time ever where the approach has been used (unfortunately, there are citations 14, 15, 67-69 about similar [?] work, lines 475, 476), or is it just for the first time for a water body in Brazil (lines 472, 473)?

Authors’ reply: qPCR assays for quantification of the sxtA gene have been designed and published by other researchers, as we report in the Discussion. However, when we tested the published primer sets for the sxtA gene with template DNA from selected cyanobacterial strains and DNA from Brazilian water samples, we repeatedly got multiple amplicons, showing that the published primers were not specific to sxtA in our samples (shown for various STX-producing strains in Figure S1). We tried to change some of the qPCR conditions to solve this problem, but we did not succeed. That is the reason we designed new sxtA primers that had a high sensitivity to target the sxtA gene in the present water samples.

The sxtA primers used in our study were designed from cyanobacteria that are commonly found in freshwater in Brazil. We used nucleotide sequences of the saxitoxin synthetase gene sxtA in the selected strains Dolichospermum circinale, Aphanizomenon flos-aquae and Raphidiopsis raciborskii to design the sxtA primers.

The reason for the unsuccessful amplification using the published primers is unknown, but geographical variations among sxtA genes have recently been suggested by Zupancic et al. (2021). This new information and the reference is now added in the Discussion.

 

2) The other story is about the particular features found in the Brazilian water reservoir (Itupararanga reservoir). They appear to be not of general interest. They may weaken the first story about the saxitoxin because it is mostly more descriptive and likely too detailed to understand the prevailing situation where the molecular approach has been tested. The data mainly concerning measurements of local variables from different sites of the reservoir should be presented only in second place and in a much-reduced way compared to present, i.e., in a Supplementary file or separate publication dealing with local/regional data from water bodies. Therefore, the section "Results 3.1 Environmental variables" (lines 221 – 254) and Table 1 should be better moved to Supplementary information. The Results section should start with Phytoplankton composition.

Authors’ reply: We agree that presentation of the different environmental variables might not be of general interest to international readers, yet the data are important to our study. The section on environmental data is now moved from section 3.1 to “Supplementary Material”, and a short summary of the environmental data is now found in the first part of section 3.4 (previously section 3.5). This appears to be an appropriate introduction to the section on “Linkage between environmental variables, cyanobacteria and cyanotoxins”.

 

3) Figure 2 is concise and well done; it provides an impressive overview of the observed phytoplankton community and abundance of cyanobacteria. Why is it useful to compare biomass with biovolume? My non-expert impression is that biovolume is more affected by errors/wrong estimates than biomass?

Authors’ reply: Biovolume is an estimation of biomass as stated by Hillebrand et al. (1999): "the estimation of biovolume has several positive aspects, which make it a recommended measure of biomass, such as: high taxonomic resolution, relatively cheap and simple to apply method and restricted to few sources of errors that are often under researcher control". We present phytoplankton abundance in units of biovolume instead of biomass to not confuse these data with the biomass data estimated from pigments by HPLC. Thus, biovolume data originate from microscopy, while biomass was determined by pigments. It should be mentioned that phytoplankton biomass can be reliably estimated from biovolume data according to Wetzel and Likens (2000), where 1 mm³ L^-1 = 1 mg L^-1. Thus, our biovolume results can be understood as biomass results.

 

4) In the Discussion section, 4.1 Environmental variables and cyanobacteria should be moved move to 2nd place. The Discussion should start with the more general information discussed in lines 570-574, followed by 4.2. qPCR assay, cyanotoxin, and cyanobacteria.

Authors’ reply: We think there is a logic flow in the present Discussion. First we present environmental control and possibly also biological control of cyanobacteria in the reservoir, followed by development of the sxtA assay and discussion of sxtA data. Next, relations between sxtA and SXT relative to environmental variables is discussed, followed by a short presentation of the other toxin, microcystin, and finally we present options for better monitoring of toxic cyanobacteria. However, we have moved the first lines in section 4.5 to section 4.1, since they introduce the problem of cyanobacteria in waters used for human purposes.

 

5) There is a minor issue with taxonomy. In line 260, Chlorophyceae is not correct. Green algae, in general, should replace it. The green algae comprise the phyla Chlorophyta (with the classes Chlorophyceae and Trebouxiophyceae) and Streptophyta (with the class Zygnematophyceae). Also, the usage of "Chlamydophyceae" is not correct; the Authors´ probably mean the order Chlamydomonadales, which is within the class Chlorophyceae. Similarly, replace Bacillariophyceae with the more general term "diatoms" (Diatomea); the species recorded from the reservoir may, in fact, belong to several classes and not only the Bacillariophyceae.

Authors’ reply: We have fixed the minor issue with taxonomy according to reviewer’s comment. You can see the changes in manuscript file (text and Figure 2) and Supplementary Material (Tables S2 and S4).

Round 2

Reviewer 1 Report

The authors have adequately addressed my concerns with the latest draft.

Reviewer 2 Report

The authors' replies are convincing and the points raised by the reviewer are dealt with adequately. The manuscript is ready for publication now.