Microbial Ecology Metagenomic analysis of zinc-surface-associated marine biofilms --Manuscript Draft-Manuscript Number:
MECO-D-18-00621R1
Full Title:
Metagenomic analysis of zinc-surface-associated marine biofilms
Article Type:
Original Article
Section/Category:
Environmental Microbiology
Order of Authors:
Wei DING Weipeng Zhang Nabeel Mannalamkunnath Alikunhi Zenon Batang Bite Pei Ruojun Wang Lianguo Chen Abdulaziz Al-Suwailem Pei-Yuan Qian
Corresponding Author:
Wei DING HONG KONG University of science and technology CHINA
Corresponding Author Secondary Information: Corresponding Author's Institution:
HONG KONG University of science and technology
Corresponding Author's Secondary Institution: First Author:
Wei DING
First Author Secondary Information: Order of Authors Secondary Information: Funding Information: Abstract:
Biofilms are a significant source of marine biofouling. Marine biofilm communities are established when microorganisms adhere to immersed surfaces. Despite the microbeinhibiting effect of zinc surfaces, microbes can still attach to the surface and form biofilms. However, the diversity of biofilm-forming microbes that can attach to zinc surfaces and their common functional features remain elusive. Here, by analyzing 9,000,000 16S rRNA gene amplicon sequences and 270 Gb of metagenomic data, we comprehensively explored the taxa and functions related to biofilm formation in subtidal zones of the Red Sea. A clear difference was observed between the biofilm and adjacent seawater microbial communities in terms of the taxonomic structure at phylum and genus levels, and a huge number of genera were only present in the biofilms. Saturated alpha-diversity curves suggested the existence of more than 14,000 operational taxonomic units in one biofilm sample, which is much higher than previous estimates. Remarkably, the biofilms contained abundant and diverse transposase genes, which were localized along microbial chromosomal segments and co-existed with genes related to metal ion transport and resistance. Genomic analyses of two cyanobacterial strains that were abundant in the biofilms revealed a variety of metal ion transporters and transposases. Our analyses revealed the high diversity of biofilmforming microbes that can attach to zinc surfaces and the ubiquitous role of transposase genes in microbial adaptation to toxic metal surfaces.
Response to Reviewers:
Reviewer #1: The manuscript "Enrichment of Transposase Genes in Zinc-Surface-
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Associated Biofilms" by Ding et al. reported the taxa and functions related to biofilm formation of zinc surfaces in subtidal zones of the Red Sea using combined 16S rRNA gene amplicons and metagenomic data. Microbial communities of biofilms showed higher diversity and novel compared to those of the adjacent seawater samples. Interesting, transposase genes were enriched in biofilms and these genes were supposed to co-exist with genes related to metal ion transport and resistance, which were further proved by the cyanobacterial genomes. The conclusions seem to make sense, but several revisions are needed. Major concerns: 1.Title of the manuscript only mentioned enrichment of transposase genes, but the manuscript included research of whole microbial communities and metagenomes. It seems that the title could not summarize the whole research. Please change it. Reply: The authors are very grateful for the positive feedback provided by the reviewer. The title has been changed to ‘Metagenomic analysis of zinc-surface associated marine biofilms’. 2.The research target is biofilms formed on zinc surface. However, zinc was not mentioned in the abstract. Furthermore, the author should clearly explain why they choose zinc to do their study in the introduction part. Reply: It is a good suggestion. We chose zinc to do this study because zinc is toxic to microorganisms and thus has anti-fouling activity. Zinc significantly reduces biofilm biomass in both intertidal and subtidal areas (Mayer-Pinto et al., 2011). There are several studies focus on the effect of zinc on biofilm formation and growth (Sabate et al., 2007; Mayer-Pinto et al., 2011; Wu et al., 2013). However, insights of these studies are limited to the influence of zinc on the amount of biomass in different stages of biofilms under a variety of environments, and the gene-level mechanism of zinc affecting biofilm formation in the marine environment remains largely unexplored. Relevant changes have been made in the revision (please see lines 65-76 in the marked-up manuscript). References: Mayer-Pinto M, Coleman RA, Underwood AJ, Tolhurst TJ (2011) Effects of zinc on microalgal biofilms in intertidal and subtidal habitats. Biofouling 27:721-727. Sabater S, Guasch H, Ricart M, Romani A, Vidal G, Klunder C, Schmitt-Jansen M (2007) Monitoring the effect of chemicals on biological communities. The biofilm as an interface. Anal Bioanal Chem 387:1425-1434. Wu C, Labrie J, Tremblay YDN, Haine D, Mourez D, Jacques M (2013) Zinc as an agent for the prevention of biofilm formation by pathogenic bacteria. J Appl Microbiol 115:30-40. 3.The taxonomy classification should be improved in L220-237. Bacillariophyta belongs to Eukaryotes but not bacteria. The cyanobacterial reads in Figure 1 may contain Chloroplast that should be removed from microbial composition analysis. Highly abundant cyanobacteria could be revealed in figure 1B, please explain it. Respectively, Figure 1 should be revised. Reply: Bacillariophyta and Chloroplast have been removed from Figure 1b. Other genera belonged to Eukaryotes were also removed, including Bangiophyceae, Chlorarachniophyceae, Cryptomonadaceae, and Streptophyta. Accordingly, the abundance of Cyanobacteria/Chloroplast has been revised in Figure 1a. At the phylum level, Cyanobacteria were enriched in biofilms in comparison with seawater microbial communities. However, a large number of microbes could not be classified down to genus level, and thus it’s highly possible that many members of Cyanobacteria were grouped into the ‘unclassified’ at genus level. Please see the revised Figure 1 in the marked-up manuscript.
4.L325-336. The comparison of genes encoding transposases and 16S rRNA genes were meaningless. There were no inescapable connections between them. Please Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
removed. Reply: The comparison of transposases genes and 16S rRNA genes has been removed. Accordingly, we have revised Figure 4. 5.L372-392. Please explain why the author chose cyanobacterial for further analysis but not other high abundant microbes. Reply: Cyanobacteria showed high abundance and high diversity in biofilms, and most cyanobacteria could not be classified down to genus level, suggesting that they are likely to be novel lineages. In addition, cyanobacteria are thought to be found in all oceans around the world, and they have important ecological functions such as providing oxygen and nutrients for other marine organisms. Please see lines 401-407 in the marked-up manuscript. Minor comments: 1.L57. I cannot understand the meaning of "substratum". Please make sure it is right. Reply: Substratum has been changed to ‘substrate’. Please see line 55 in the markedup manuscript. 2.L127-128. The length of v2-v4 region is usually more than 500bp, and a 2*300bp sequencing is needed. I am not sure that the 2x250bp sequencing is enough for your research. Reply: In this study, the V3V4 region (forward primer: 5’-CCTACGGGNGGCWGCAG3’; reverse primer: 5’-GACTACHVGGGTATCTAATCC-3’) was used. The length of V3V4 region in this study is around 460 bp.
Reviewer #2: The study brings new information about the diversity of microorganisms in the biofilm of zinc surfaces and the role of transposase genes in microbial adaptation on toxic surfaces. A good study, however there are some issues that need to be elucidated. Abstract: Why analyze only two Cyanobacteria strains? Reply: The authors are grateful for the constructive comments. Cyanobacteria showed high abundance and high diversity in biofilms, and most cyanobacteria could not be classified down to genus level, suggesting that they are likely to be novel lineages. In addition, cyanobacteria are thought to be found in all oceans around the world, and they have important ecological functions such as providing oxygen and nutrients for other marine organisms. “We analyzed two cyanobacterial genomes to confirm the functional features related to biofilm formation on the zinc surfaces. These two genomes were selected because cyanobacteria showed high abundance and high diversity in the biofilms, and most cyanobacteria could not be classified down to the genus level, suggesting that they are likely to be novel lineages. In addition, cyanobacteria are thought to be found in all oceans around the world and have important ecological functions such as providing oxygen and nutrients for other marine organisms.” Please see lines 401-407 in the marked-up manuscript. If there is a difference between biofilm and water how can we allege lines 33-34? Reply: Regarding the diversity of microbes and their biogeographical patterns, the principle of ‘everything is everywhere, but, the environment selects’ implies that all microorganisms are distributed worldwide, while in a given environmental setting most species can be latent present (Beijerinck, 1913; Baas Becking, 1934; De Wit and Bouvier, 2006). Microorganisms in water can attach easily to solid surfaces and form biofilms, which explains why microbial biofilms are found almost everywhere (Salta et Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
al., 2013; Dang et al., 2016). Here, microbes (e.g. cyanobacteria) in seawater were selected during the process of biofilm assembly, and some rare species in seawater could proliferate after forming biofilms. However, as also suggested by other reviewers, we have revised the overstatement. Please see lines 37-39 in the marked-up manuscript. References: Beijerinck MW (1913) De infusies en de ontdekking der backteriën. Jaarboek van de Koninklijke Akademie voor Wetenschappen. Amsterdam: Müller. (Preprinted in Verzamelde geschriften van M.W. Beijerinck, vijfde deel 119–140. Delft, 2921). Baas Becking LGM (1934) Geobiologie of inleiding tot de milieukunde. The Hague, the Netherlands: W.P. Van Stockum & Zoon (in Dutch) De Wit R Bouvier T (2016) ‘Everything is everywhere, but, the environment selects’; what did Baas Becking and Beijerinck really say? Environ Microbiol 8(4):755-758. Salta M, Wharton JA, Blache Y, Stokes KR, Briand JF (2013) Marine biofilms on artificial surfaces: structure and dynamics. Environ Microbiol 15(11):2879-2893. Dang H, Lovell CR (2016) Microbial surface colonization and biofilm development in marine environments. Microbiol Mol Biol Rev 80(1):91-138. Abstract session has a lack of information to lead to its conclusions. Reply: We have revised the conclusion. “Our analyses revealed the high diversity of biofilm-forming microbes that can attach to zinc surfaces and the ubiquitous role of transposase genes in microbial adaptation to toxic metal surfaces.” Please see lines 37-39 in the marked-up manuscript. Introduction is appropriate however it gives the impression that it was assembled from cutouts of a larger text. Needs revision! Could, also, bring the hypothesis of the study. Reply: We have revised the Introduction by adding more research about the effect of zinc on biofilms, and a hypothesis. “As one type of the widely studied substrates, metal surfaces are known to be toxic to microorganisms. In particular, zinc can significantly reduce biofilm biomass in both intertidal and subtidal habitats; laboratory experiments also suggest that low concentrations of zinc inhibit biofilm formation of Actinobacillus pleuropneumoniae.” Please see lines 65-69 in the marked-up manuscript. “We hypothesized that clear difference could be observed between the biofilm and seawater microbial communities, and microbes may adopt certain ubiquitous mechanisms to facilitate their attachment to the zinc surfaces. To test this hypothesis, we need to (a) make a high-resolution estimate of the number of species that can form biofilms on zinc surfaces; and (b) explore the functional basis for microbes to form biofilms on zinc surfaces.” Please see lines 85-91 in the marked-up manuscript. Methods: According to literature, biofilms could take up to 60 days to reach its maximum development stage. Why was the time of 30 days chosen for the study? Reply: The duration required for biofilm to develop to “mature stage” is dependent of environmental condition. For instance, in Hong Kong water, it only takes about 12 days for biofilm to reach its maximal stage during the summer. Because biofilm can have several developmental stages, in the present study we focus on the effect of zinc on the early stage of biofilm development. The biofilms developed on the panels for 30 days have accumulated enough biomass for metagenomic sequencing. Did the zinc panels have rugosity? It is known that the roughness of the substrate influences its colonization by microorganisms, this factor was considered? Reply: Yes, the zinc panels are rough. The roughness of the substrate surface may have influence on biofilm formation. In a previous study, Gharechahi et al. (2012) Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
explored the different effect of surface roughness and material composition on biofilm formation; they showed that both the surface structure and composition of the underlying surface have an impact on the initial bacterial adhesion. Moreover, Morgan et al. (2001) investigated that after incubated with bacteria for 4 h, surface roughness appeared to have no effect on the number of adherent bacteria. Therefore, the roughness of surface affected the very initial stages of biofilm formation. In the present study, subtidal biofilms were developed on the panels for 30 days, and thus it is long enough to minimize the effect caused by surface roughness. Moreover, in the present study we focus on the possible mechanisms that facilitate microbial adaptation to the zinc, and thus didn’t consider much about the surface roughness. References: Gharechahi M, Moosavi H, Forghani M (2012) Effect of surface roughness and materials composition on biofilm formation. J Biomater Nanobiotechnol 3:541-546. Morgan TD, Wilson M (2001) The effects of surface roughness and type of denture acrylic on biofilm formation by Streptococcus oralis in a constant depth film fermentor. J Appl Microbiol 91:47-53. Cell density was calculated only for biofilm samples, what was the aim of this information? Reply: The cell density result was used to show an overall view of biomass in zinc surface associated biofilms. Based on other reviewers’ comments, we think the cell density analysis doesn’t add much to the major finding, and thus plan to delete this part to make the results more clear and compact. There is a mathematical formula to calculate cell density that consider a correction to the use of formaldehyde, it has been used? In epifluorescence analysis we have to consider that. Reply: It is a good suggestion. However, we could not find this normalized formula in other published papers. As mentioned above, we have removed the cell density result. Results and Discussion Lines 239-240: Would this information invalidate some of the data obtained? If zinc panels already had been coated with microbial cells, this would change the biofilm development. How panels were treated before they were submerged? Reply: Sorry for the misleading sentences. The zinc panels were washed with autoclaved distill water to make sure no microbes on it before submerging them in seawater at a depth of 3 m for 30 days. “The zinc panels were washed with autoclaved distill water to make sure no microbes on it before submerging them in seawater at a depth of 3 m for 30 days.” Please see lines 97-98 in the marked-up manuscript. Lines 242-244: Not sure that can state this if the process was already started in the zinc panels. The information in previous lines should be clarified thus puts in question all the following discussion. Reply: Sorry for the confusing expression. The panels are almost sterile before deployed into seawater. Line 251: Which mechanisms? Reply: Here we want to say that in addition to species sorting, there must be other mechanisms underlying the observed different microbial community structures between biofilm and seawater. We have made it more clear in the revised manuscript. Please see lines 270-273 in the marked-up manuscript. Lines 278-280: What could explain, then, the differences between this and previous studies? The substrate?
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Reply: The major reason for the different microbial diversities between this and previous studies should be sequencing depth. We believe that the microbial diversity observed in previous studies were not adequately captured, given the high microbial diversity of marine biofilms. And this certainly adds the value of the present study. “Taken together, these findings indicate that the microbial diversity in marine biofilms had previously been underestimated, probably due to inadequate sequencing depth, and that a huge number of marine microbes have the ability to attach to metal surfaces and form biofilms.” Please see lines 306-309 in the marked-up manuscript.
Lines 281-282: Sampling points were pretty close so not sure if can consider geographic location here. Reply: It is a good suggestion. We have revised the overstatement. “On the other hand, the beta-diversity analysis suggests that the variation in geographic location have sampling sites has little effect on the taxonomic composition of the biofilms.” Please see lines 303 in the marked-up manuscript.
Reviewer #3: In this manuscript, the authors investigate biofilms on zinc-coated platinum surfaces. The approaches they use include metagenomic sequencing and 16S rRNA gene amplicon sequencing for the biofilm and the sea water samples. Their results show that the transposase genes are highly enriched in biofilms formed on the zinc surface. The transposase genes are also seen to be related with metal transport and resistance genes. This implies that these genes could be important factors which enable microorganism to overcome metal toxicity and form biofilms on toxic metal surfaces. Overall this is a well-designed and executed study and I only have the following relatively minor suggestions/comments for the authors to consider: Line 88: Where were the biofilms scrapped? At the collection point? In the lab? Reply: The authors are very grateful for the positive feedback provided by the reviewer. The biofilms were scrapped at the collection point, stored in DNA extraction buffer, and then delivered to lab with dry ice. Line 91: Here, it is mentioned that sea water was filtered through 0.22 µm membrane. In the supplementary information, it is said that the sea water was filtered through 0.1 µm membrane. Which one is it? Also, the nature of membrane has not been mentioned, nor the volume of seawater sampled. Reply: Sorry for the mistake. Here 0.1 µm membranes were used. The membranes we used to filter seawater were hydrophilic polytetrafluoroethylene membranes, which show an excellent collection efficiency (>93%) for bacteria and viruses (Burton et al. 2006). Please see lines 103-105 in the marked-up manuscript. Reference: Burton NC, Grinshpun SA, Reponen T (2006) Physical collection efficiency of filter materials for bacteria and viruses. Ann Occup Hyg 51:143-151. http://doi.org/10.1093/annhyg/mel073 Line 100: Concentration of DAPI used? Reply: In this study, the final concentration of DAPI (5 ng/µl) was used, as described in our previous publication (Tian et al. 2014). However, considering comments from other reviewers, we have removed the cell density part. Reference: Tian RM, Wang Y, Bougouffa S, Gao ZM, Cai L, Zhang WP, Bajic V, Qian PY (2014) Effect of copper treatment on the composition and function of the bacterial community Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
in the sponge Haliclona cymaeformis. MBio 5(6):e01980-14. Cell density calculation: I am surprised that they could count the number of cells accurately from a scrapped biofilm at 40X magnification. Would be helpful if a representative image of the DAPI stained filter could be provided in the main text or SI. Reply: Based on other reviewers’ comments, we finally decided to remove the cell density part, because this part is problematic and doesn’t add much to the major conclusion. Line 137: What is the basis of picking up 500,000 reads? Usually, to even out the sequencing depth, all the samples are rarefied to the minimum number of reads obtained for a particular sample in the data set. Else its just throwing away data and not analyzing the whole thing. The lowest number of reads in the data set belongs to sample 2-Water-1 which is 788,539 sequences. By analyzing 500,000 sequences, we are basically losing out on 288539 sequences from each of the other samples. Though in this case, the sequencing looks adequate enough even at 500,000 samples, but as a practice I think one should analyze as much data as possible. Reply: Good suggestion! We realized that rarefying the samples to the equal size may not be a good way, as also indicated in a related study (McMurdie et al., 2014). Thus we have analyzed the 16S data again by skipping the step of data size normalization. The result is included in the revised manuscript. “In addition, to analyze as many sequences as possible, OTU classification based on all the 16S rRNA gene sequences (without data size normalization) was performed, which revealed similar taxonomic compositions (Fig. S3), suggesting a minor impact of data size normalization on profiling the taxonomic structure of the biofilm and seawater microbial communities.” Please see Figure S3 and lines 253-257 in the marked-up manuscript. Reference: McMurdie PJ, Holmes S (2014) Waste not, want not: why rarefying microbiome data is inadmissible. PLoS Comput Biol 10(4):e1003531.
Line 142: How the Singletons were removed? Reply: Singletons (OTUs with only one read) in a given sample were classified into ‘minor’. Line 241: I do not exactly get what the author is trying to imply here. Zinc surface already had a microbial layer before they started their experiment? Reply: This sentence was misleading. We have removed this sentence. The zinc panels were washed with autoclaved distill water to make sure no microbes on it before deployed into the sea. Please see lines 97-98 in the marked-up manuscript. Line 252: "... but are able to proliferate during biofilm assembly." - need to rephrase. Reply: These mechanisms are likely related to rare species that are undetectable in seawater but are able to proliferate after biofilm formation. Please see lines 270-273 in the marked-up manuscript. Line 264: Surprisingly, more than 18,000 OTUs were observed in each biofilm sample on average. Figure 2 shows only one biofilm sample (Biofilm_1) has 18,000 OTUs at > 450000 reads. Then how could the average be more than 18,000 OTUs? Reply: Sorry for the mistakes. The OTUs have been revised by removing certain eukaryotic species. Here we revised it to ‘the samples on average have 16,348 OTUs’. Please see line 285 in the marked-up manuscript. Line 314: Accounted for 2.27%. Seems like an incomplete sentence. Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Reply: A higher percentage of transposases than free-living microorganisms was also identified in the biofilms from an acid mine drainage, where transposases accounted for 2.27% of the total metagenomic reads. Please see lines 334-337 in the marked-up manuscript. Line 322: According to their reason, they say microbes that attach to surface already possess transposase genes. In that case, if normalized by biomass, there should be no enrichment of transposase genes observed. If that is so, is their conclusion still valid? Reply: This is a good question. In the present study, we analyzed the relative abundance, rather than absolute abundance of different microbes in communities. Actually it is difficult to perform comparison of absolute abundance of the microbes. In these sentences, we want to say that before the biofilm-forming microbes attach to the zinc surface, the transposase genes are already in their genomes; but microbes with transposase genes are enriched by biofilm formation. Thus the enrichment of transposases genes is not related to the biomass of microbial communities. Line 341: Extracellular DNA from biofilms usually maps on to complete genome of the biofilm forming organisms. Its size has been shown to be roughly the same as genomic DNA. So, it might be that it is still in the extracellular DNA portion. The original paper which came up with this theory, justified the conclusion based on elimination of other sources such as viral genomes and plasmids (ref: ISME J 2009, 3, 1420). They have not proven that small extracellular fragments contain transposase genes. They took these 3 choices, and by data analysis proved that viral genomes and plasmids were not a contributing factor. Thus, left with small extracellular DNA fragments from biofilms. The cited literature here also hypothesizes that small extracellular DNA fragments may possibly be carriers. Not experimentally proven. Reply: We have revised the overstatement. “The identification of transposase genes in long contigs also indicates that transposase genes are likely located in the chromosomal DNA molecules rather than in small DNA fragments (e.g., plasmids and viral genomes). This finding is generally in agreement with the hypothesis in a study of deep-sea biofilms, which excluded viral genomes and plasmids as carrier of transposase genes. However, within a biofilm community, it’s difficult to determine whether the transposase genes are localized in the chromosomes of living bacterial cells or in extracellular DNA.” Please see lines 364-368 in the marked-up manuscript. Fig 4A: Orange in figure corresponds to 16S rRNA gene contigs, blue corresponds to contigs with transposase. In the legend its mentioned wrongly. Reply: Thanks for the comments. But based on comments from other reviewers, we have already removed the 16S rRNA result in this figure.
Reviewer #4: Great work. Congratulations. Reply: The authors are grateful for the very positive feedback of the reviewer.
Reviewer #5: General comment: Essentially, this manuscript describes the sessile and planktonic microbial communities in seawater nearby or attached to submerged zinc surfaces, and finds an interesting variety and preponderance of transposase genes and metal-related functions in zincassociated biofilms. The paper is well written, the methodology appears to be Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
appropriate, and the conclusions are discussed with respect to the data. The figures are attractive and clear. Minor comments to address: PAGE 14 LINE 373: Can you please clarify that these Cyanobacteria genomes were the result of your genome discovery, perhaps by adding "Cyanobacteria" to PAGE 7 LINE 192. Reply: The authors are grateful for the positive feedback provided by the reviewer. We have made the statement more clear: “The completeness and potential contamination of the two cyanobacterial genomes recovered in the present study were estimated using the software CheckM. The two genomes are summarized in Table S3.” Please see lines 207 and 209 in the marked-up manuscript. PAGE 14 LINE 390: Can you comment on the presence of any kind of mobile elements that might be associated with these Cyanobacteria transposases (and therefore the metal-resistance genes)? In other words, are these transposes orphaned or are they associated with IS elements or conjugative transposons? Reply: We have revised Figure S8 (shown below). In cyanobacterial genomes, we found both orphaned transposase (Fig. S8b) and transposase associated with insertion elements or conjugative transposons (Fig. S8c). For instance, PF13612 (transposase with DDE domain), IRL (inverted repeats left), and IRR (inverted repeats right) form a transposon; this transposon, K07480 (insertion gene insB), and PF13586 (transposase with DDE_Tn_1_2 domain) are adjacent to each other and probably form a complex transposon. Suggestions: PAGE 1 LINE 16 "Biofilm communities are established when microorganisms adhere to surfaces immersed in marine environments". Suggest change to "Marine biofilm communities are established when microorganisms adhere to immersed surfaces." Reply: Revised according to the comment. Please see lines 17-18 in the marked-up manuscript. PAGE 1 LINE 21 "Here by analyzing 9,000,000 16S rRNA gene amplicon sequences and 270 Gb of metagenomic data, we comprehensively explored the taxa and functions related to biofilm formation in subtidal zones of the Red Sea." Suggest adding a comma after "Here". Reply: Revised according to the comment. Please see line 23 in the marked-up manuscript. PAGE 2 LINE 34 "Latently" The exact meaning that you wish to convey with this word is unclear. Do you mean transiently? Reply: As suggested by other reviewers, we have removed this sentence. PAGE 2 LINE 43 Suggest changing "Microbes residing in marine biofilms secrete cues for larval settlement by recruiting fouling invertebrates such as the barnacle Balanus Amphitrite and the polychaete Hydroides elegans." To "Microbes residing in marine biofilms secrete recruiting cues for larval settlement of fouling invertebrates such as the barnacle Balanus Amphitrite and the polychaete Hydroides elegans." Also, perhaps give an example of what these cues are? Reply: Moreover, microbes residing in marine biofilms secrete inductive cues (e.g., isobutyl methylxanthine) for larval settlement of fouling invertebrates (e.g., the polychaete Hydroides elegans), resulting in macro-fouling problems. Please see lines 46-48 in the marked-up manuscript. PAGE 3 LINE 80: Typo: "Basis"? and delete the "the" before "zinc surfaces".
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Reply: Revised according to the comment. The sentence has been changed to “explore the functional mechanisms for microbes to form biofilms on zinc surfaces.” Please see line 90 in the marked-up manuscript.
Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Revised Manuscript
Click here to access/download;Manuscript;ManuscriptMarkedUp.docx
Click here to view linked References
1
Enrichment of Transposase Genes in Zinc-Surface-Associated Biofilms
2
Metagenomic analysis of zinc-surface-associated marine biofilms
3 4
Wei Dinga, Weipeng Zhangb, Nabeel Mannalamkunnath Alikunhic, Zenon Batangc,
5
Bite Peib, Ruojun Wangb, Lianguo Chena, Abdulaziz Al-Suwailemc, Pei-Yuan Qiana,b,*
6
a
7
Kong, China
8
b
9
Hong Kong, China
Division of Life Science, Hong Kong University of Science and Technology, Hong
Department of Ocean Science, Hong Kong University of Science and Technology,
10
c
11
Saudi Arabia
12
*Email:
[email protected]
King Abdullah University of Science and Technology, Thuwal, The Kingdom of
13 14
Keywords: Marine biofilm, Zinc panel, Transposase, Metagenome
15 16
Abstract
17
Biofilms are a significant source of marine biofouling. Marine biofilm communities
18
are established when microorganisms adhere to immersed surfaces. Biofilm
19
communities are established when microorganisms adhere to surfaces immersed in
20
marine environments. Despite the microbe-inhibiting effect of zinc surfaces, microbes
21
can still attach to the surface and form biofilms. However, the diversity of biofilm-
22
forming microbes that can attach to zinc surfaces and their common functional
23
features remain elusive. Here, by analyzing 9,000,000 16S rRNA gene amplicon
24
sequences and 270 Gb of metagenomic data, we comprehensively explored the taxa
25
and functions related to biofilm formation in subtidal zones of the Red Sea. A clear
26
difference was observed between the biofilm and adjacent seawater microbial
27
communities in terms of the taxonomic structure at phylum and genus levels, and a
28
huge number of genera were only present in the biofilms. Saturated alpha-diversity 1
29
curves suggested the existence of more than 14,000 operational taxonomic units in
30
one biofilm sample, which is much higher than previous estimates. Remarkably, the
31
biofilms contained abundant and diverse transposase genes, which were localized
32
along microbial chromosomal segments and co-existed with genes related to metal ion
33
transport and resistance. Genomic analyses of two cyanobacterial strains that were
34
abundant in the biofilms revealed a variety of metal ion transporters and transposases.
35
Taken together, we propose that a large proportion of biofilm-forming microbes are
36
only latent present in seawater before attaching to metal surfaces, and these microbes
37
often carry a huge number of transposase genes. Our analyses revealed the high
38
diversity of biofilm-forming microbes that can attach to zinc surfaces and the
39
ubiquitous role of transposase genes in microbial adaptation to toxic metal surfaces.
40 41
Introduction
42
In the marine environment, biofilms are sessile communities formed by microbes that
43
secrete extracellular polymeric substances upon adhering to a surface. These microbes
44
mainly include bacteria and archaea. Biofilm can directly affect the condition of man-
45
made surfaces in aquatic environments [1-3] and cause serious microbial fouling
46
problems [4-6]. Moreover, microbes residing in marine biofilms secrete inductive
47
cues (e.g., isobutyl methylxanthine) for larval settlement of fouling invertebrates (e.g.,
48
the polychaete Hydroides elegans), resulting in macro-fouling problems [7-8].
49
Microbes residing in marine biofilms secrete cues for larval settlement by recruiting
50
fouling invertebrates such as the barnacle Balanus Amphitrite and the polychaete
51
Hydroides elegans. Biofilms can cover almost all wetted surfaces, causing serious
52
microbial and macro-fouling problems and leading to billions of dollars in economic
53
losses each year [4-6].
54 55
With the aim of preventing biofilm formation, the effect of substrate substratum type
56
on biofilm community composition has been examined for a variety of materials [92
57
12]. For example, in a marine cold seep system, biofilms are formed on different
58
materials by microbes with different taxonomic affiliations, pointing to a ‘species
59
sorting’ process [9]. However, the effect of the substrate seems stronger in the initial
60
stage than in the later stage of biofilm formation [10-12]. A study of subtidal
61
biofilms has similarly found that the substratum type has a strong influence on the
62
biofilm composition during initial developmental stages and that this effect
63
diminishes as the biofilm ages [7]. Such discoveries indicate a ‘convergence’
64
process; that is, the effect of substratum type is reduced as the biofilm forms and
65
develops. As one type of the widely studied substrates, metal surfaces are known to
66
be toxic to microorganisms. In particular, zinc can significantly reduce biofilm
67
biomass in both intertidal and subtidal habitats [13]; laboratory experiments also
68
suggest that low concentrations of zinc inhibit biofilm formation of Actinobacillus
69
pleuropneumoniae [14]. However, it is frequently observed that bacterial biofilms can
70
withstand the effects of toxic metals better than planktonic cultures of the same
71
species [15]; the presence of a small population of persister cells may contribute to the
72
time-dependent tolerance of biofilms to metal cations [16]. Thus, in the aquatic
73
environment, microbes can attach to both metal and nonmetal surfaces and form
74
biofilms. These findings have implied the influence of metal surfaces on behavior of
75
microbes and suggested the existence of ubiquitous mechanisms governing biofilm
76
formation on metal surfaces in nature.
77 78
Although a quantity of studies have reported the interactions between metal surfaces
79
and the biomass of microbial biofilms and between metal surfaces and the
80
composition of microbial biofilms, the diversity of biofilm-forming microbes that can
81
attach to metal surfaces and their common functional features remain elusive. Recent
82
advances in sequencing technology have facilitated unprecedented analyses of the
83
ocean microbiome. In the present study, using both 16S rRNA gene amplicon
84
sequencing and metagenomic sequencing, we examined biofilms formed on 3
85
submerged zinc-coated panels in three subtidal locations of the Red Sea. We
86
hypothesized that clear difference could be observed between the biofilm and
87
seawater microbial communities, and microbes may adopt certain ubiquitous
88
mechanisms to facilitate their attachment to the zinc surfaces. To test this
89
hypothesis, we need to (a) make a high-resolution estimate of the number of species
90
that can form biofilms on zinc surfaces; and (b) explore the functional mechanisms
91
for microbes to form biofilms on zinc surfaces.
92 93
Methods
94
Sampling information
95
The three locations in the central Red Sea for biofilm collection are shown in Fig. S1.
96
Biofilms were formed on the surface of zinc-coated aluminum wire panels (Fig. S2).
97
The zinc panels were washed with autoclaved distill water to make sure no microbes
98
on it before submerging them in seawater at a depth of 3 m for 30 days. Three biofilm
99
biological replicates and three adjacent seawater biological replicates were collected
100
from each location, resulting in 18 samples. Biofilms were scraped off the surfaces
101
using autoclaved cotton tips after loosely attached particles were washed off with
102
0.22-μm-filtered 0.1-μm-filtered seawater. The adjacent seawater samples were
103
filtered through a 0.22-μm 0.1-μm membrane. The membranes used to filter seawater
104
were hydrophilic polytetrafluoroethylene membranes, which show an excellent
105
collection efficiency (>93%) for bacteria and viruses [17]. All of the samples were
106
stored in DNA storage buffer (10 mM Tris-HCl; 0.5 mM EDTA, pH 8.0), shipped to
107
laboratory with dry ice, and then kept at -80°C until DNA extraction.
108 109
Cell Density Calculation
110
The number of bacteria was enumerated under a microscope using fluorescent dye
111
staining. Briefly, bacterial cells were scraped off a panel surface and re-suspended in 1
112
ml of PBS-buffered formaldehyde (2%). The re-suspended cells were filtered through 4
113
a polycarbonate membrane filter (0.22-μm pore size) that was placed on a filter
114
manifold, before 25 μl of 4',6-diamidino-2-phenylindole (DAPI) were added to the
115
filtration well to stain the microbial cells. After incubating for 20 minutes in the dark,
116
the filter was placed on a microscope slide and observed under a confocal laser
117
scanning microscope (Carl Zeiss, Jena, Germany) at 40 × magnification. For bacterial
118
enumeration, 10 fields were counted for each sample. Cell density was then
119
transformed to the number of cells per cm2 based on the ratio between the area of the
120
microscopic field and that of the panel.
121 122
DNA Extraction
123
Vortex was performed to release the microbial cells from cotton tips (in the case of
124
biofilms) and from filter membranes (in the case of seawater samples). All of the
125
samples were subjected to centrifugation at 10,000 rpm for one minute and the
126
supernatant was discarded. The bacterial cells were treated with 10 mg/ml lysozyme
127
and further with 20 mg/ml proteinase K. DNA extraction was performed using a
128
microbial genomic DNA extraction kit (TIANGEN Biotech, Beijing, China)
129
following the manufacturer’s protocol.
130 131
16S rRNA Gene Amplification, Sequencing, and Analyses
132
The quality of DNA extracts was determined using 1% agarose gel electrophoresis
133
before Polymerase chain reaction (PCR). The hypervariable V3V4 region (forward
134
primer:
135
GACTACHVGGGTATCTAATCC-3’) [18] of prokaryotic 16S rRNA genes was used
136
to amplify DNA from biofilm and seawater microbial communities. PCR was
137
performed in a thermal cycler (Bio-Rad, USA) using the Phusion DNA polymerase
138
(Finnzymes Oy, Espoo, Finland). The PCR program consisted of initial denaturation
139
at 98°C for 30 s; 26 cycles of first 98°C for 10 s, then 55°C for 10 s, and then 72°C
140
for 15 s; and finally 72°C for 5 min. The PCR products were purified before library
5’-CCTACGGGNGGCWGCAG-
5
3’;
reverse
primer:
5’-
141
construction and sequencing at Novogene (Beijing, China) on the HiSeq 2500
142
platform. The read length was 250 bp and each pair of reads had a 50 bp overlapping
143
region.
144 145
The 16S sequencing data are summarized in Table S1. The 16S rRNA gene amplicon
146
data were analyzed using the software packages QIIME (version 1.91) [19] and
147
USEARCH [20]. The paired-end reads were subjected to quality control using the
148
NGS QC Toolkit (version 2.0) [21] and were merged using the ‘fastq_mergepairs’
149
script implemented in USEARCH. The ‘fastq_filter’ and ‘uchime_denovo’ commands
150
in USEARCH were used to remove low-quality reads and chimeras. To normalize the
151
uneven sequencing depth, 500,000 filtered reads for each sample were picked up.
152
OTUs were classified de novo from the pooled reads at 97% sequence similarity using
153
the ‘pick_otus’ script and then representative sequences were recovered using the
154
‘pick_rep_set’ script in QIIME. Normalized reads were then mapped to the
155
representative sequences using the ‘otutab’ command in USERACH to generate an
156
OTU table. Singletons from the OTU table were removed during taxonomic profiling
157
(they were classified as ‘minor’ during taxonomic profiling). Taxonomy was classified
158
using the RDP classifier. Rarefaction curves and alpha-diversity analyses (observed
159
OTUs, Shannon diversity, and Simpson diversity) were performed using scripts in
160
QIIME. Beta-diversity was revealed by the principal coordinates analysis (PCoA)
161
based on the Bray-Curtis distances calculated using the software PAST [22].
162
Additionally, profiling of the taxonomic community structures was performed based
163
on the 16S rRNA gene amplicons without data size normalization.
164 165
Metagenomic Sequencing and Analyses
166
Metagenomic sequencing was conducted at Novogene (Beijing, China) on the HiSeq
167
2500 platform, after construction of 350 bp short insert libraries. In total, more than
168
270 Gb of metagenomic DNA sequences (150 bp read length and paired-end 6
169
sequencing) were generated for the 18 samples. Reads information of the high-
170
throughput sequencing data are summarized in Table S2. The paired-end Illumina
171
reads were subjected to quality control using the NGS QC toolkit [21] on a local
172
server, which removes reads containing more than 30% of low quality bases.
173 174
Metagenomes of the three biofilm or seawater replicates per sampling location were
175
assembled using MEGAHIT (version 1.0) [23]. A series of k-mers ranging from 71 to
176
111 were used during the assembly process. The total number of contigs per
177
assemblage ranged from 1,047,671 to 2,819,648. Reads from the three replicates were
178
mapped to the assembled contigs individually to calculate the coverage of the contigs.
179
Open reading frames (ORFs) in the assembled Illumina reads were predicted using
180
Prodigal (version 2.6.3) [24].
181 182
Functional community structure was analyzed using BLASTp, E
