Edited by: Fernando Poyatos-Jimenez, University of Seville, Spain
Reviewed by: Zeynep Basaran Bundur, Özyeğin University, Turkey; Rodorico Giorgi, Consorzio Interuniversitario Per Lo Sviluppo Dei Sistemi A Grande Interfase, Italy; Guadalupe Pinar, University of Natural Resources and Life Sciences Vienna, Austria; Julio Romero-Noguera, Seville University, Spain
This article was submitted to Microbiotechnology, a section of the journal Frontiers in Microbiology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
To overcome the limitations of traditional conservation treatments used for protection and consolidation of stone and lime mortars and plasters, mostly based on polymers or alkoxysilanes, a novel treatment based on the activation of indigenous carbonatogenic bacteria has been recently proposed and applied both in the laboratory and
Stone and lime plaster deterioration is one of the most serious problems affecting historical structures and sculptures all over the world (
Microorganisms are able to cause several types of damage on monument surfaces, including biophysical, biochemical, and aesthetic biodeterioration, which may occur simultaneously or separately (see reviews by
The conservation of such historic and culturally important artworks typically involves the application of consolidating agents that in many cases do not provide long-lasting efficacy and induce further damage due to pore blocking as well as severe aesthetic alterations (
The first step to understand the relationship between microorganisms and environment, as well as its potential response during a bio-conservation treatment, is to identify the bacteria present in the cultural artwork. Next generation sequencing approaches have been developed to study the complexity of microbial communities in a wide range of environments, allowing in-depth studies of environmental samples (
This study was performed at the Maya site of Copan, Honduras (
The Maya site of Copan:
For the testing of the bioconsolidation treatment based on the application of the patented M-3P nutritional solution, two types of substrates (tuff stone and lime plaster) were considered in this study:
Stone block (sample CI) located in the base section of a vertical wall of a Late Classic Temple (Structure 18) made up of a buff-colored carved tuff stone, showing extensive deterioration and material loss due to scaling (
Decontextualized stone block (sample CPN) originally located on the ground next to the Copan Acropolis (
Block of Late Classic lime plaster floor (sample MC) from Structure 12 (AD ∼800). After excavation, it was stored at the Copan Sculpture Museum storage area. This plaster (
Note that this biotreatment application was a trial aimed at gauging its potential effectiveness under the particular exposure conditions in this tropical region. In order to minimize possible negative side-effects in case of treatment failure (i.e., possible activation of harmful bacteria leading to substrate acidification or discoloration), we have restricted the extent of its application for minimum impact. The extent of the testing and sampling was further restricted taking into account the value and uniqueness of this Maya site. Therefore, the three selected substrates were chosen considering their representativeness regarding the different materials and exposure conditions at the site. The first substrate (stone at Structure 18) is representative of the Copan tuff stones of different structures at the Copan Acropolis; the decontextualized stone block is representative of the excavated stone blocks and sculptures currently located at the Copan Sculpture Museum and storage area; and the plaster is representative of interior lime plaster floors and walls in structures at the Copan Acropolis, as well as plaster pieces located at the Copan Sculpture Museum and storage area.
The treatment application procedure is described in detail in
Both
Genomic DNA was extracted from solid samples collected aseptically (using sterile tweezers, and sterile Eppendorf tubes) from all three samples, both before treatment application (CI_CONT, CPN_CONT, and MC_CONT: taken at time 0, before the application of M-3P nutritional solution) and 3 months after treatment application (CI_ TREATED, CPN_TREATED, and MC_TREATED). Three replicates of each sample were performed. After collection at the site, each sample/replicate (∼0.5 g) was aseptically placed in a 2-mL sterile screw-cap tube containing glass beads. One mL of lysis buffer (100 mM Tris-HCl [pH 8.0], 100 mM EDTA [pH 8.0], 100 mM NaCl, 1% polyvinylpyrrolidone [PVP], and 2% SDS), 24 μL freshly made lysozyme (10 mg/mL), and 2 μL proteinase K (20 mg/mL) were added to each tube. The tubes were vigorously shaken for 2 min in vortex followed by mechanical lysis of the cells performed twice using a FastPrep® FP120 (at a speed of 5.5 m/s for 45 s) with intermittent cooling on ice for 5 min. The tubes were incubated at 37°C for 30 min and then at 60°C for another 30 min, and subsequently centrifuged at 14,000 ×
High-throughput amplicon sequencing using 250 bp paired-end sequencing chemistry (MiSeq Illumina) was performed. Total DNA of each sample was amplified targeting the hypervariable V3-V4 regions by using the 16S rRNA gene primers 341F and 785R (
Demultiplexing of all libraries was carried out for each sequencing lane using the Illumina bcl2fastq. After quality controlling and combining using PandaSeq (
Species richness was measured through the use of alpha-diversity metrics (Chao1, Shannon diversity index, and observed species) in Explicet and R softwares. Beta-diversity, the similarity between the identities of taxa, and their abundances in each sample were assessed using Bray-Curtis distances (weighted UniFrac distances) measured in QIIME and PAST3 v. 3.18 and the output was visualized by means of principal coordinate analysis (PCoA). A heatmap was constructed for the visualization of specific differences in community composition using the heatmap.2 function in the R gplots v2.11.0 package on log-normalized abundance data. At the genus level, the heatmap included only taxa at ≥1.5% relative abundance in all samples/replicates. Relative abundance graphs were constructed representing relative microbial abundance averages of three biological replicates. Additionally, similarity of percentages analysis (SIMPER) was performed using PAST3 software. Network analyses were conducted in the R environment using the VEGAN package and only strong Pearson’s correlations (ρ > 0.8 or ρ < −0.8) were considered. Network visualization and modularization were carried out on the interactive platform of Cytoscape (
All raw sequences used in this study are available in the sequence read archive (SRA) at NCBI database under the SRA accession number
A total of 910,782 bacterial 16S rRNA gene sequences were recovered for all samples. They were distributed in a mean of 63,420 for CI_CONT, 40,835 for CI_TREATED, 46,518 for CPN_CONT, 219,251 for CPN_TREATED, 4,990 for MC_CONT, and 46,108 for MC_TREATED. These sequences were used for community analyses by QIIME and OTUs were assigned by clustering sequences with over 97% sequence identity. A number of 1,165 OTUs were identified, indicating high microbial diversity.
Alpha-diversity analysis revealed no significant differences in bacterial richness of the studied samples regardless of the metric used (
Inference statistics at genus level of the different samples of the Maya archeological site.
SAMPLES | Taxa_richness | Simpson_1-D | Shannon_H | Pielou’s evenness | Fisher’s Alpha | ACE | CHAO1 | Goods coverage index |
CI_CONT | 1561 | 0.94 | 3.77 | 0.51 | 291.81 | 137.3 | 139.1 | 0.999 |
CI_TREATED | 496 | 0.54 | 1.8 | 0.29 | 79.8 | 105.9 | 107.4 | 0.999 |
CPN_CONT | 2592 | 0.96 | 5.10 | 0.65 | 603.1 | 315.1 | 314.4 | 0.998 |
CPN_TREATED | 4619 | 0.91 | 3.61 | 0.43 | 850.1 | 219.6 | 223.1 | 0.998 |
MC_CONT | 239 | 0.98 | 4.65 | 0.85 | 47.9 | 85.3 | 85.9 | 0.998 |
MC_TREATED | 1108 | 0.93 | 3.67 | 0.53 | 211.2 | 141 | 138.1 | 0.999 |
Beta-diversity revealed significant differences in bacterial community structure and abundance between the studied samples before and after the M-3P treatment. The samples formed three distinct clusters, in accordance with the structure type, localization and substrate type, on the principal coordinate analysis (PCoA = multidimensional scaling, MDS) plot of the bacterial community composition (using OTU abundance) of all tested samples (
Principal coordinates analysis (PCoA) showing the relationship between the bacterial population structures of the stone and plaster samples in the Maya archeological site, untreated and treated with the M-3P nutritional solution (based on Bray–Curtis index). CI: Temple structure (Structure 18); CPN: stone block treated in the LACEM laboratory; and MC: lime plaster block treated in the LACEM laboratory.
Among the 30 phyla determined in all three samples,
Taxonomic distribution of the bacterial community at phylum level in the stone and plaster samples of the Maya archeological site. Averages of relative abundance of three biological replicates from each sample are represented. Only phyla with a relative abundance of more than 0.1% are represented. CI: stone block of the Temple structure; CPN: stone block treated at the LACEM laboratory; and MC: lime plaster block treated at the LACEM laboratory.
In the stone sample treated
At genus level, identification of the bacterial groups revealed the presence of 485 different genera in the bacterial communities of all samples (
Taxonomic distribution of the bacterial community at genus level in the stone and plaster samples of the Maya archeological site. Averages of relative abundance of three biological replicates from each sample are represented. Only genera with a relative abundance of more than 0.2% are represented. CI: stone block of the Temple structure; CPN: stone block treated at the LACEM laboratory; and MC: lime plaster block treated at the LACEM laboratory.
To further identify the similarity in abundance among the bacterial communities, a heatmap was produced based on the relative abundance of the genera with an average abundance of >1.5% in at least one sample, which were defined as dominant. Abundance of 35 major genera present in all the samples was illustrated (
Heatmap based on the relative abundance of genera with an average abundance of >1.5% in at least one sample/replicate. Averages of relative abundance of three biological replicates from each sample are represented.
Similarity of percentages analysis (SIMPER) was used to determine the relative contribution of each individual taxon to the dissimilarity among the three substrates (CI, CPN, and MC). The average Bray–Curtis dissimilarity and the contribution of each genus to the total dissimilarity between communities of treated and untreated samples were calculated, and the top major genera responsible for the microbial community difference (>98% contribution to cumulative dissimilarity) are summarized in
SIMPER analysis of bacterial community dissimilarity (>98% of contribution to cumulative dissimilarity) of the three untreated Maya substrates (CI_CONT: stone sample
Taxon | Avg dissimilarity (%) | Contribution to dissimilarity (%) | Cumulative dissimilarity (%) | Mean abundance (%) | ||
Mean CI_CONT | Mean CPN_CONT | Mean MC_CONT | ||||
18.24 | 21.33 | 21.33 | 51.3 | 0.494 | 0.282 | |
7.931 | 9.276 | 30.61 | 22.2 | 0.153 | 0 | |
6.092 | 7.125 | 37.73 | 0.00152 | 21.3 | 0 | |
5.333 | 6.236 | 43.97 | 1.1 | 17.6 | 9.63 | |
4.981 | 5.826 | 49.79 | 13.9 | 0.497 | 0 | |
3.496 | 4.088 | 53.88 | 0.00442 | 7.18 | 9.4 | |
2.374 | 2.776 | 56.66 | 0 | 0.00134 | 6.65 | |
1.519 | 1.776 | 58.43 | 0.198 | 2.35 | 4.45 | |
1.421 | 1.662 | 60.09 | 0.00656 | 4.09 | 2.84 | |
1.381 | 1.615 | 61.71 | 0.00694 | 0.0196 | 3.87 | |
1.354 | 1.583 | 63.29 | 0.000829 | 4.74 | 0 | |
1.31 | 1.532 | 64.82 | 0.22 | 2.25 | 3.02 | |
1.178 | 1.377 | 66.2 | 0 | 0 | 3.3 | |
Unclassified_JG35-K1-AG5 | 1.133 | 1.325 | 67.53 | 0 | 0 | 3.17 |
0.9831 | 1.15 | 68.68 | 0.512 | 0.493 | 3.06 | |
0.9415 | 1.101 | 69.78 | 0.54 | 3.58 | 0.761 | |
Unclassified_ |
0.8125 | 0.9502 | 70.73 | 0.0348 | 0.0135 | 2.27 |
Unclassified_ |
0.6897 | 0.8066 | 71.53 | 0.118 | 0.312 | 2.05 |
0.6463 | 0.7558 | 72.29 | 0.101 | 0 | 1.75 | |
Unclassified_ |
0.6423 | 0.7512 | 73.04 | 0 | 0.00293 | 1.8 |
0.6238 | 0.7295 | 73.77 | 0 | 0.619 | 1.68 | |
0.6036 | 0.7059 | 74.48 | 0.00809 | 0.0369 | 1.7 | |
0.6028 | 0.705 | 75.18 | 2.24 | 0.415 | 1.93 | |
0.5787 | 0.6768 | 75.86 | 0 | 0.809 | 1.62 | |
0.5766 | 0.6744 | 76.53 | 0.0258 | 0.0181 | 1.63 | |
0.5615 | 0.6567 | 77.19 | 0.0143 | 1.96 | 0 | |
0.5386 | 0.6299 | 77.82 | 0.106 | 1.91 | 0.184 | |
0.5352 | 0.6259 | 78.45 | 0.165 | 1.81 | 1.06 | |
0.4541 | 0.5311 | 78.98 | 0.0403 | 1.53 | 0.424 |
SIMPER analysis of bacterial community dissimilarity (>98% of contribution to cumulative dissimilarity) of the three treated Maya substrates (CI_TREAT: stone sample treated
Taxon | Avg dissimilarity (%) | Contribution to dissimilarity (%) | Cumulative dissimilarity (%) | Mean abundance (%) | ||
Mean CI_Treat | Mean CPN_Treat | Mean MC_Treat | ||||
29.09 | 35.74 | 35.74 | 87.30 | 0.02 | 0.01 | |
17.22 | 21.16 | 56.90 | 0.06 | 51.70 | 32.80 | |
6.94 | 8.53 | 65.43 | 0.01 | 6.82 | 20.80 | |
5.09 | 6.26 | 71.69 | 0.01 | 1.84 | 15.30 | |
Unclassified_ |
4.16 | 5.11 | 76.80 | 0.59 | 13.10 | 2.69 |
2.88 | 3.54 | 80.35 | 0.15 | 8.80 | 0.37 | |
2.52 | 3.10 | 83.44 | 0.01 | 5.19 | 7.58 | |
1.94 | 2.39 | 85.83 | 0.22 | 0.20 | 5.98 | |
1.57 | 1.93 | 87.76 | 0.02 | 4.73 | 0.84 | |
1.43 | 1.76 | 89.52 | 4.30 | 0.00 | 0.00 | |
Unclassified_ |
1.10 | 1.35 | 90.87 | 0.02 | 0.02 | 3.30 |
Unclassified_ |
0.68 | 0.83 | 91.70 | 0.01 | 0.36 | 2.03 |
Unclassified_ |
0.65 | 0.80 | 92.50 | 0.00 | 0.08 | 1.96 |
0.62 | 0.77 | 93.27 | 1.89 | 0.13 | 0.03 |
To comprehensively understand the interaction effects between bacteria detected before and after the application of the nutritional solution, a network of significant co-occurrence and co-exclusion relationships among genera was constructed on the basis of strong Pearson correlation matrix (ρ > 0.8 or ρ < −0.8) (
Network analysis revealing co-occurrence patterns among bacterial taxa in the different samples of the Maya archeological site. The nodes and edges are colored according to Betweenness Centrality. Only taxa with a relative abundance of more than 0.5% in at least one sample were used. Each connection represents strong correlations based on Pearson’s correlation coefficient (ρ of >0.8 and ρ of > –0.8). The thickness of each line is proportional to the significance of the interaction (ρ-value), and the size of the circle (the node) is proportional to the number of connections, i.e., the degree, of bacterial genera.
High-throughput sequencing results in our study revealed a rich diversity of bacterial communities on the deteriorated stone and plaster samples of the Maya archeological site. We identified a total of 161 genera in the CI_Cont, 309 in CPN_Cont, and 141 in MC_Cont. Association networks constructed to the genus level showed mainly positive correlation patterns with strong and complex connections, as well as some negative correlations with loose and simple connections. These interactions between microbial taxa in stone and plaster artworks indicated significant co-occurrence patterns helping to decipher the relationships between community members. Most of these genera belong to
Generally, microbial communities in archeological sites have been reported to have distinct diversity patterns under different environmental conditions (
In general, the microbial communities living in the archeological sites or historical buildings all over the world are linked to environmental conditions (
The diversity of the bacterial communities in samples CPN (stone) and MC (mortar) was almost twice as large as compared with sample CI (stone in Structure 18). Possibly, the direct contact with the soil microbiota of the ground had a significant impact on the bacterial population in the former samples. Remarkably, the degree of similarity was higher between the stone sample CPN and mortar sample MC than between the two volcanic tuff stone samples, indicating that the original sample location was of greater importance than substrate characteristics (composition, mineralogy and textural features). Additionally,
Finally, members of the phylum
Regardless of treatment location (i.e.,
Our results show that the composition of the original bacterial population has an important influence on the bacterial population developed after the consolidation treatment. In the case of the stone sample from Structure 18 (CI),
In samples CPN and MC,
Remarkably, members of the phylum
In this study, we showed for the first time the enormous impact of the bioconsolidation methodology based on the application of a sterile nutritive solution (M-3P) on the indigenous bacteria present in stone and plaster at the Maya archeological site of Copan. A detailed characterization of the bacterial population evolution revealed that the bioconsolidation treatment induced a significant increase in beneficial indigenous carbonatogenic bacteria and a concomitant suppression of potentially damaging
Furthermore, the relative safety and ease of application of the treatment proposed here as compared to conventional biotreatments using bacteria inoculums are worth highlighting. The former simply consists of spraying the patented nutritional solution onto the degraded substrate surface, allowing for large-scale applications without posing any significant risk to operator or environment. Admittedly, the biotreatment is labor intensive (i.e., relatively large number of applications are required). However, overall material-related treatment costs are comparable to those of conventional consolidation treatments.
All raw sequences used in this study are available in the sequence read archive (SRA) at NCBI database under the SRA accession number
CR-N and MTG-M conceived the concept and led this project. CR-N and FJ designed the experimental setup. CR-N, KE, and ER-A set up and managed the treatment and the sample collections. FJ performed all laboratory works, analyzed the data, and created the graphs and figures. FJ, KE, and CR-N were major contributors in writing the manuscript with critical input from MTG-M, and ER-A. All authors read and approved the final version of the manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We thank William Fash and Barbara Fash as well as the team at the Maya Sculpture Conservation Laboratory (LACEM) in Copan for their help and continuous support, as well as for providing access to this Maya archeological site and their contribution to the sampling and treatment application process. We also thank Alberto de Tagle and Nieves Valentin for their help and support during the development of this study.
The Supplementary Material for this article can be found online at: