Microbiome of seventh-century old Parsurameswara stone monument of India and role of desiccation-tolerant cyanobacterium Lyngbya corticicola on its biodeterioration

Abstract The Parsurameswara stone monument, built in the seventh century, is one of the oldest stone monuments in Odisha, India. Metagenomic analysis of the biological crust samples collected from the stone monument revealed 17 phyla in the microbiome, with Proteobacteria being the most dominant phylum, followed by cyanobacteria. Eight cyanobacteria were isolated. Lyngbya corticicola was the dominant cyanobacterium in all crust samples and could tolerate six months of desiccation in vitro. With six months of desiccation, chlorophyll-a decreased; however, carotenoid and cellular carbohydrate contents of this organism increased in the desiccated state. Resistance to desiccation, high carotenoid content, and effective trehalose biosynthesis in this cyanobacterium provide a distinct advantage over other microbiomes. Comparative metabolic profiles of the biological crust and L. corticicola show strongly corrosive organic acids such as dichloroacetic acid, which might be responsible for the biocorrosion of stone monuments.


Introduction
Historical monuments are a significant source of sociocultural value around the world.However, over time, the deteriorative and degradative action of physical, chemical, and biological agents leads to biodeterioration.The severity of biodeterioration is primarily determined by factors such as the extent and type of microbial growth, the nature of the colonized material, and the level of pollution present (Gabriele et al. 2023).Cyanobacteria are critical in deteriorating modern and historical architecture (Sterflinger and Piñar 2013).Although some studies have considered both cyanobacteria and green algae as major biofilm-forming phototrophs, comparatively few have examined their influence on the deterioration of historical stone monuments.Microbiomes of stone monuments of tropical climates are unique because they experience high temperatures, desiccation and high UV irradiation.
The microbiome inhabiting the biological crust on stone surfaces is often not culturable.Metagenomic studies are culture-independent approaches to understanding the composition and functional potential of complex mixed microbial communities of different environmental conditions.However, metagenomic studies on the biological crusts of many stone monuments are unknown.In tropical climates, cyanobacteria and other algae act as primary sources of nutrients for other organisms and promote the adhesion of microorganisms to stones.This causes cracking of the underlying materials through the microbial influence on the substratum.Lithobiotic cyanobacteria and other microorganisms lead to additional stone weathering through geochemical activity and can transform organic molecules into different forms (Yadav et al. 2021).Thus, the biochemical activity of microbes is a serious problem for preserving cultural heritage sites (Scheerer et al. 2009).Biochemical reactions with the substratum can cause discoloration in appearance, alteration of porosity and vapour/moisture diffusivity in and out of the stone, acid production, solubilization, and mobility of ions and salts when interacting with the surrounding environment.Cyanobacteria can resist extreme environmental stressors such as extreme desiccation, high light intensity, elevated temperatures and salinity (Albertano and Urz� ı 1999) by adapting to stressful environments.They can produce stress-specific cellular proteins called adaptation proteins, secondary metabolites (Burja et al. 2003;Gupta et al. 2013), enzymes, such as superoxide dismutase (SOD), catalase (CAT), and guaiacol peroxidase (GPX) (Latifi et al. 2009), and non-enzymatic substances, such as carotenoids and glutathione reductase, which synthesize and excrete extracellular polymeric substances (EPS) (Latifi et al. 2009).Extracellular polymeric substances help cyanobacterial strains survive in exposed terrestrial environments, but they can also cause stone decay and enhance biodeterioration (Adhikary and Sahu 1998;Rossi et al. 2012).However, the metabolic profile of crust cyanobacteria in stone monuments is unknown.Mass spectrometry (MS) is an effective technique used to identify and analyse metabolites in intricate biological systems.In the present study, we conducted a comparative metabolomic profile of crust samples of Parsurameswara stone monuments and one dominant cyanobacterium, Lyngbya corticicola, using LC-ESI-MS/MS and attempted to understand the role of its metabolites in the biodeterioration of Parsurameswara stone monuments.
Many stone monuments, including Parsurameswara, are open to the air and are readily colonized by different microorganisms.However, the microbiome of these ancient stone monuments remains unknown.Microbiomes are the leading cause of the deterioration of stone monuments and cause severe destruction over time (De Leo and Urz� ı 2015;Salvadori and Municchia 2016).Understanding the microbiome colonizing the surface of culturally important monuments and their metabolic profile can be crucial in preserving them.By studying the microorganisms present, we can gain insights into their potential effects on the deterioration and develop strategies to mitigate them.Visible colonization of cyanobacterial mats can be observed in many of the sculptures and carvings of monuments.However, the diversity of microbiomes colonizing these stone monuments has not yet been studied.Since many of the microbiomes in stone monuments are not culturable, we used a metagenomic approach to study them.In addition, we were able to culture a dominant cyanobacterium, Lyngbya corticicola, which we found in all crust samples, study its ecophysiology under desiccated conditions, compare its metabolic profile with biological crust samples, and try to understand its role in the biodeterioration of this monument.

Nature of the stone monument, location of collection site, and sample collection procedure
The Parsurameswara monument is situated between 20 � 14 0 35.27 00 N latitude and 85 � 50 0 20.57 00E longitude in Bhubaneswar city of the Indian state of Odisha (Figure 1).The monument measures 3.011 � 2.97 m from the inside, 6.01 � 6.4 m from the outside and has a height of 12.26 m.The Parsurameswara Monument is a unique architectural, historic stone monument of India built in the seventh century.It is a Nagara-style Hindu temple comprising three main types of stones.Khondalite was primarily used for the monument body and most sculptures; high-quality chlorite was used for a few large sculptures and carvings, and the interior core was constructed using laterite stones.We used nondestructive double-layered adhesive tapes (25 � 50 mm) to collect cyanobacterial crust samples, transferred them to pre-sterilized screw cap bottles, and analysed them in the laboratory.Crust samples from various parts of the stone monument covering all surfaces were collected.A Garmin Oregon 550 GPS was used to record the coordinates of the sampling site, while to measure the temperature on the surface of the stone, an infrared thermometer was used.Additionally, the relative humidity was measured using a hygrometer.

Isolation, culture, and identification of the cyanobacteria from crust samples
To isolate cyanobacteria from the collected crust sample, a pinch of crust from the adhesive tape was soaked in sterile distilled water for 12-48 h.Crust samples were then observed using a stereo-zoom microscope.The green part of the crust was aseptically transferred to agar plates (1.2% w v -1 agar in BG-11 � and BG-11 media) and incubated under fluorescent light (95 lmol m −2 s −1 light intensity) with 14/10 h light/dark cycle at 25 ± 1 � C for 15-20 days.The cyanobacteria that appeared in the culture were repeatedly sub-cultured to obtain a pure axenic culture of cyanobacteria.The purity of the axenic culture was checked by repeated subcultures on nutrient agar plates with 0.5% glucose (w v -1 ), incubated at 35 � C for 24 h to observe contaminating bacteria if any.The pure culture was then deposited in the Visva-Bharati Culture Collection of Algae (VBCCA), affiliated with the World Federation of Culture Collection (WDCM 931), and assigned the strain number.Standard monographs have been used to identify documented cyanobacteria (Desikachary 1959;Komarek and Anagnostidis 2008;Komarek 2013).

Metagenomic DNA extraction, 16s rDNA amplicon library generation, and sequencing
Metagenomic DNA extraction, 16s rDNA amplicon library preparation, and sequencing were performed sequentially to obtain the metagenomic profile of the biological crust of the Parsurameswara stone monument.The total genomic DNA of the crust samples collected from various parts of the Parsurameswara stone monuments was prepared using the DNeasy Power Soil Kit (Qiagen, Maryland, USA) according to the manufacturer's protocol.The DNA extracted from the samples was analysed in NanoDrop (2000, Thermo Fisher Scientific, Inc., Pittsburgh, PA, USA), and the isolated environmental DNA quality was checked on 0.8% agarose gel before being taken for PCR amplification.For metagenomic analysis, environmental DNA was amplified in the V3-V4 hypervariable region of the 16S rDNA gene of bacteria and archaea using the primers, Prokaryote V3-Forward (5 0 -CCTACGGGNBGCASCAG-3 0 ) and Prokaryote V4-Reverse (5 0 -GACTACNVGGGTATCTAATCC-3 0 ).Extracted environmental DNA (40 ng) was amplified using 10 pM of each primer.Amplicons from each sample were purified using Ampure beads to remove unused primers.An additional eight cycles of PCR were performed using Illumina (Hayward, CA, USA) barcoded adapters to prepare sequencing libraries.Libraries were purified using Ampure beads and quantified using the QubitV R dsDNA High Sensitivity assay kit (Invitrogen, CA, USA).The amplicon libraries were checked on an Agilent DNA1000 chip (Agilent Technologies, CA, USA) after purification by 1� AMpureXP beads on Bioanalyzer2100 and quantified using Qubit dsDNA HS Assay kit by Qubit Fluorometer 2.0.Sequencing was performed using an Illumina Miseq with a 2x300PE v3 sequencing kit.Metagenomic sequence analysis was conducted using QIIME 2 (Caporaso et al. 2010), an open-source microbiome data science platform.Initially, the raw sequence (in fastq format) was visualized using FastQC version 0.11.9 (Andrews 2017).
PCR primer sequence of the V3 region of the respective samples was removed from the sequencing data using cutadapt (v.1.8.3).Sequence data in fastQ format was trimmed using sickle (v.1.33)in paired-end mode with default settings (Earl et al. 2011).Forward and reverse reads were merged into a single amplicon read using fastq-join, allowing fragments with a length of 200-250 bp for V3.A minimum phred score of 38 was used for downstream analysis.The entire cleaned sequence data was concatenated into a single file.Denoising the reads to correct it and getting the amplicon sequence variant (ASVs) with a 16S reference (88% OTUs from Green genes 13_8) as a positive filter was carried out.The references are only used for evaluating the likelihood of each sequence being 16S through local alignment using SortMeRNA, with a permissive e-value; however, the reference is not used to characterize the sequences.Assigning taxonomy to ASVs was conducted against SILVA_132_16S_V3_V4 database.Here 'classify-sklearn' was used, which is a pre-fitted sklearn-based taxonomy classifier.Filtration of the feature table was used to remove rare variants and contaminants (chloroplast and mitochondria) and unclassified ASVs as obtained after taxonomy assignment.Further, stacked bar chat pf taxa relative abundances were generated.Feature/OTU heatmap at the genus level was generated.The SEPP method is used to place short sequences into a reference phylogenetic tree.Generating a rarefaction curve can determine whether or not sufficient sequencing is done.QUIME was used to determine the Chao 1 index, alpha diversity and rarefaction curve representing community diversity.The alpha and beta diversity metrics of the sample were also analysed from the data.The metagenomic sequences were deposited in NCBI GenBank under Bioproject PRJNA924525 with the accession number SAMN32759396 and can be accessed with the following link https://www.ncbi.nlm.nih.gov/sra/PRJNA924525.

Screening of desiccation-tolerant cyanobacteria
The isolated cyanobacteria were subjected to an experiment inside the silica gel desiccator to identify the most desiccation-tolerant cyanobacteria.The eight isolated cyanobacteria were Calothrix clavatoides, Nostoc linckia, Phormidium ambiguum, Arthrospira plantensis, Gloeocapsa sanguine, Aphanothece microscopia, Scytonema schmidtii and Lyngbya corticicola.During the exponential growth phase, eight cyanobacterial species were harvested by centrifugation.They were allowed to dry with 50% relative humidity inside a Klenz Flo (Mumbai, India) laminar airflow under a stream of sterile air.Respective air-dried cyanobacteria were placed inside a silica gel desiccator and incubated under fluorescent light (95 lmol m −2 s −1 light intensity) with 14/10 h light/dark cycle at 25 ± 1 � C and 5% relative humidity for 30 days.At every three-day interval, 2 mg of desiccated cyanobacterial samples were removed from the desiccator, and their growth in terms of chlorophyll-a was measured.Chlorophyll-a of the respective cultures of dried cyanobacteria (2 mg) in the zero-day desiccation experiment was used as a control.Among the eight cyanobacteria, the one with the maximum growth in the desiccated state was selected for further experimentation.

Pigment and biochemical analysis of desiccationtolerant cyanobacterium Lyngbya corticicola
Among the eight isolated cyanobacteria from the biological crust samples of the Parsurameswara stone monument, Lyngbya corticicola grows well and survived 30 days under the desiccation condition with 5% relative humidity and was selected for the further long-term experiment of 6 months.This experiment was conducted in six different desiccators and 15 mg air-dried Lyngbya corticicola cells were incubated inside the desiccator and kept under fluorescent light (95 lmol m −2 s −1 light intensity) with 14/10 h light/dark cycle at 25 ± 1 � C with 5% relative humidity.All the six desiccators was appropriately labelled, and individual air-dried samples weighing 5 mg were placed separately in each desiccator for subsequent analysis of pigments, cellular carbohydrates and cell proteins.Every month, a 5 mg portion of desiccated Lyngbya corticicola cells, which had been set aside separately at the start of the experiment, was extracted for analysis of pigment composition, cellular carbohydrates, and cell proteins.For pigment analysis, 5 mg of desiccated cells were extracted in 90% methanol (v v -1 ).The absorption spectra were measured in a Shimadzu (Kyoto, Japan) UV-1800 UV-visible double beam spectrophotometer in the wavelength range from 250 to 700 nm using quartz cuvettes.The quantity of chlorophyll-a was estimated using the extinction coefficient of Mackinney (1941), and the amount of carotenoids was determined following Davis (1976).Cellular carbohydrate contents of the cells after different months of desiccation were estimated using anthrone reagent as described by Herbert et al. (1969).The quantities of proteins in the cells after different months of desiccation were also determined following Lowry et al. (1951).

Extraction of metabolites from the crust and L. corticicola for metabolic profiling
For metabolite extraction, 50 ml of DCM: MeOH (1:1) was added to 1 g of biological crust collected from the Parsurameswara stone monument and lyophilized L. corticicola cells, and incubated for 20 min followed by sonicated in the water bath for 10 min.Further, the process was repeated for methanol, ethyl acetate and acetone extraction.The entire extract was filtered using Whatman grade 1 filter paper, and any extra solvent was evaporated and lyophilized through rotary evaporation at 42 � C.

Liquid chromatography-electrospray ionization tandem mass spectrometry analysis
LC-ESI-MS/MS analysis of the metabolites extracted from the biological crust and L. corticicola was performed with a Waters ACQUITY UPLC M-Class System (Waters Corp., Milford, MA, USA).It is equipped with Xevo G2-XS Q-Tof MS (Waters Corp.) via an electrospray ionization system (ESI).The lyophilized metabolites were dissolved in 1 ml of methanol (1 mg ml −1 ) and filtered using a 0.2 lm PTFE membrane filter (Phenomenex, Torrance, CA, USA).Twenty ml of the sample were then diluted to 60 ml with 0.1% formic acid before being injected into the Acquity UPLC-BEH C18 column (pore size 130 A 0 , particle size 1.7 mm, inner diameter 2.1 mm � length 100 mm) with a trapping time of 3 min.The mobile phase consisted of (A) water with 0.1% formic acid (FA) and (B) acetonitrile (ACN) with 0.1% FA.The flow rate was kept at 30 ml min −1 .The total run time of 11 min was set as follows: 0 to 2 min equilibrium, 2 to 11 min gradients from 0% to 90% of mobile phase B and from 90% to 0% of mobile phase A, additional 2 min washing with 0.1% FA.After passing through the LC column, the eluent was analysed using MS in the ESI (þve) platform at 3 KV.The Mass Lynx Version 4.0 software (Waters Corp.) was used to record the data-dependent acquisition.A TIC chromatogram was generated, and the data produced were analysed using Progenisi Qi (Waters, Milford, MA, USA) software.The generated chromatogram and retrieved data of the whole metabolites profile of both biological crust and L. corticicola were further compared.The acquired m/z of the detected compounds were analysed and compared with the ChemSpider database (Pence and Williams 2010) and the standard cyanobacterial database, CyanometDB (Jones et al. 2021), to get the metabolic profile.Based on the abundance/intensity (the number of hits of a particular ion (m/z) received at that time (RT) in the detector) the increase or decrease in specific metabolites in crust samples compared to the pure culture of L. corticicola was calculated.

Statistical analysis
All the data obtained from the desiccation experiment for the measurement of chlorophyll-a, carotenoids, cell proteins and carbohydrates were analysed using Microsoft Excel (Microsoft, Washington, USA).The standard error of the mean was calculated with error bars.
We also isolated cyanobacteria from the crust samples of the Parsurameswara stone monument.We were able to isolate eight cyanobacteria: Calothrix clavatoides, Nostoc linckia, Phormidium ambiguum, Arthrospira plantensis, Gloeocapsa sanguine, Aphanothece microscopia, Scytonema schmidtii and Lyngbya corticicola from the biological crust samples (Figure 1e-l).Lyngbya corticicola was obtained in all the crust samples, so it could be the dominant cyanobacteria in the stone monument of Parsurameswara.Thallus of the L. corticicola upon culturing in BG 11 (þN) medium showed blue-green filaments, filaments straight, cells broader than long, 13.8-23.0lm broad, 2.3-3.5 lm long; trichomes not constricted at the cross walls; sheath lamellated, yellowish brown, 2.3-4.6 lm thick and apices rounded (Figure 1l).

Screening of the desiccated cyanobacteria
In the tropical climate of India, stone monuments experience extreme climatic conditions with temperatures of 52-60 � C with extremely dry conditions.To identify the most desiccation-tolerant cyanobacteria among the eight isolates, the cyanobacteria under investigation were subjected to air-drying and subsequent incubation within a desiccator set to a relative humidity of 5% for 30 days.The growth of these cyanobacteria was measured in terms of chlorophylla, monitored at three-day intervals throughout the experiment.The increase or decrease in chlorophyll-a (%) in the studied cyanobacteria after desiccation up to 30 days are given in Figure 3.The results shown that Lyngbya corticicola (VBCCA 03 08) was very much tolerant to desiccation.In response to desiccation, chlorophyll-a gradually decreased with an increase in desiccation duration.However, in the case of L. corticicola, a 50% reduction in chlorophyll-a was not achieved even after 30 days of desiccation.Gloeocapsa sanguine and Phormidium ambiguum were not tolerant of desiccation, and the LC 50 (50% reduction in chlorophyll-a) was achieved within the ninth day of desiccation.Scytonema schmidtii, Arthrospira plantensis, Calothrix clavatoides and Nostoc linckia were moderately tolerant to desiccation, and LC 50 was achieved after 18 days of desiccation in these cyanobacteria (Figure 3).

Pigment and biochemical composition of Lyngbya corticicola with six months of desiccation
The impact of desiccation up to six months on the pigment composition, cellular carbohydrate and cell protein is given in Figure 4. Growth in terms of chlorophyll-a decreased continuously with an increase in the duration of desiccation, and the lowest chlorophyll-a was recorded after six months of desiccation.Within six months, though, the cells survived 57% reduction in chlorophyll-a was observed.However, carotenoid content was progressively increased (Figure 4).The cellular carbohydrate content of L. corticicola increased continuously with the duration of desiccation and with six months of desiccation 158% increase in cellular carbohydrates was observed.However, cell protein decreased continuously with an increase in the duration of desiccation.

Comparative metabolic profile of biological crust and L. corticicola
To determine and compare the complete metabolite profiles of the crust sample and Lyngbya corticicola, we conducted an analysis using liquid chromatography-mass spectrometry (LC-ESI-MS/MS).The LC-MS chromatogram shown in Figure S1 (supplementary material) displayed numerous peaks according to the presence of metabolites (organic acid, mycosporin-like amino acid, sugar and other secondary metabolites).Compounds were identified by correlating the molecular ion peaks with MS fragmentation, reported by researchers and online databases.Table 1 represents the retention time (RT), m/z found in positive modes (ESIþ), and their comparative abundance of identified compounds.The public database sources that are used to retrieve molecular formulas and other chemical information are CyanoMetDB, SciFinder (http://www.cas.org/products/scifinder),PubChem (http://pubchem.ncbi.nlm.nih.gov/) and ChemSpider (http://www.chemspider.com).Alignment of obtained mass spectra with published papers (Sapse 2000;Virolleaud et al. 2006;Welker et al. 2006) resulted in the identification of three kinds of mycosporine-like amino acid (MAA); mycosporine-GABA, porphyra-334 and N-methyl palythine, organic acid; dichloroacetic acid, delta-guanidinovaleric acid and lyngbic acid and other secondary metabolites: trehalose, microginin FR9, nostoclide 2, antanapeptin A, eucapsitrione, biselyngbyaside C, carmaphycin A and aulosirazole.The comparative analysis of the biological crust of the Parsurameswara monument and the dominating desiccated cyanobacterium L. corticicola shows an interesting result.Biselyngbyaside C is absent in the crust while it is

�
Abundance/intensity ¼ the number of hits of a particular ion (m/z) received at that time (R t) in the detector.

��
Fold change: ratio between normalized metabolite abundance in L. corticicola vs normalized metabolite abundance in the biological crust.
present in L. corticicola, whereas phormidinine B is absent in L. corticicola but present in the crust.The result also shows an abundance of trehalose, N-methyl palythine, eucapsitrione and dichloroacetic acid in the biological crust.

Discussion
Microbiomes, such as bacteria, cyanobacteria and algae, inhabit culturally significant stone monuments and contribute to substantial deterioration by breaking down the stone matrix and disrupting the internal structures of the monument.Understanding the microbiome and its metabolites is crucial for formulating effective conservation strategies for stone monuments.Cyanobacteria colonies disfigure the fine sculptures and carvings of seventh-century Parsurameswara stone monuments, but the microbiome of this important monument was not known.We have studied the abundance of culturable and unculturable microorganisms and the diversity of cyanobacteria colonizing this monument.This result shows proteobacteria and cyanobacteria as the dominant bacterial phyla in the stone surfaces.Metagenomics analysis shows Arthospira, Lyngbya, Leptolyngbya, Anabaena, Scytonema, Calothrix, Trichomus, Geminocystis, Microcystis and Synechococcus as the most abundant species among cyanobacteria.However, when we culture the crust samples, Lyngbya corticicola was found in all the crust samples.This contradiction might be due to the unavailability of a sequence of L. corticicola in the NCBI Gen Bank.Recently, Skipper et al. (2022) worked on the metagenomic analysis of the bacterial microbiome of limestones of different churches and found that within the communities, Actinobacteria, Proteobacteria and Cyanobacteria were the most abundant phyla.Metagenomics of the microbiome of the Beishiku temple in North-West China was studied by Wu et al. (2022).They analysed the microbial diversity, taxon composition, and biochemical potential of the microbiome in the sandstone temple of Beishiku.Cyanobacteria support the establishment and development of heterotrophic organisms by supplying low molecular weight carboxylic acids obtained from the fixation of atmospheric carbon.
Cyanobacteria facilitate the colonization and growth of many heterotrophic organisms such as Flavobacterium, Pseudomonas, Burkholderiales, Comamonadaceae, Myxococcales, Planctomycetaceae, Rhizobiales and Sphingomonadaceae by providing small molecular weight carboxylic acids derived from atmospheric carbon fixation.Many reports of the biodeterioration of historical monuments worldwide by these microbiomes, including cyanobacteria, are available (Meng et al. 2020).In addition, organic acids such as lactic, gluconic, citric and tartaric acids excreted by the microbiome have been proven destructive to monuments (Kirtzel et al. 2020).Based on an in-depth knowledge of these processes, sustainable control strategies can be developed more effectively against the biodeterioration of stone monuments and buildings (Bao et al. 2023).Though some studies are available on colonizing microbiomes in stone monuments, most are revealed up to the class level.Among the metagenomics studies, it is described as an operational taxonomic unit (OTU), and in many studies, species or genera present in the sampled population were not determined (Schloss and Westcott 2011).In the present work, we have also done a culture-based study that shows Lyngbya corticicola as the dominant cyanobacteria isolated from all crust samples, besides seven other cyanobacteria isolated from the stone monuments of Parsurameswara.The microbiome of stone monuments varies with the nature of the substratum and environmental parameters (Keshari and Adhikary 2014).Mondal et al. (2022) documented the seasonal diversity of cyanobacteria.They reported Brasilonema sp., Aphanothece pallida, Chroococcidiopsis kashayi, Leptolyngbya boryana, Leptolyngbya polysiphoniae, Scytonema multiramosum, Westiellopsis prolifica from the stone monuments of Santiniketan and Bishnupur.The most common cyanobacteria species on monuments in Europe, America and Asia belong to the genera Gloeocapsa, Phormidium, Chroococcus and Microcoleus (Ortega-Calvo et al. 1995).The stone surface microenvironment, particularly the roughness, porosity, hygroscopicity and water absorption capability, influence the microbiome (Urzı � and Realini 1998;Prieto and Silva 2005;Miller et al. 2006).
Our study shows Lyngbya corticicola is a desiccation-tolerant cyanobacterium that can withstand six months of desiccation.We also observed that with increased desiccation duration, chlorophyll-a decreases continuously, and carotenoids increase.Carotenoids can protect cyanobacteria against photooxidative damage (Krinsky 1971) and are well known for their antioxidant activity (Karsten and Garcia-Pichel 1996).An increase in carotenoid content in response to desiccation in cyanobacterium Leptolyngbya sp. was also reported (Joshi et al. 2018).An increase in reactive oxygen species (ROS) formation occurs because of desiccation, and biosynthesis of carotenoids might occur to quench the singlet excited state oxygen.
We have observed a decline in cellular proteins with an increase in desiccation.During desiccation, the expression of genes related to photosynthesis, nitrogen metabolism and protein synthesis was found to be reduced in Anabaena PCC 7120 (Katoh et al. 2004).Certain proteins were completely lost in the cell population, as revealed by the protein profile of Anabaena PCC 7120 after desiccation (Singh et al. 2013).This explains the sensitivity of protein synthesis to desiccation stress.During cellular dehydration in cyanobacteria specific proteins related to glycolysis display decreased expression levels.In Nostoc flagelliforme, the gene responsible for phosphoglycerate kinase was observed to undergo downregulation, and a reduction in the enzymatic activity of triose-phosphate isomerase was observed in response to water loss (Wang et al. 2022).Water stress proteins (WSPs) are acidic proteins of 33.37 and 39 kDa found in Nostoc commune under desiccation stress.However, these are a substantial fraction of the cellular proteins and probably are degradation products of larger functional stress proteins (Hershkovitz et al. 1991).
We also observed a gradual increase in carbohydrates in response to desiccation.Lyngbya corticicola is surrounded by the external envelop layers formed by polysaccharides.Accumulation of sugars was observed when cyanobacteria were exposed to drought (Hershkovitz et al. 1991).Due to desiccation, carbohydrate/sheath also increases and acts as a primary colonizer, providing nutrients and moisture, which supports and stimulates the growth of heterotrophic bacteria, unicellular green algae, fungi, mosses, and gradually small plants.During desiccation and rewetting cycles, the carbohydrate produced by L. corticicola changed in volume, which might have caused the loosening of the stone joints and led to the biodeterioration of the Parsurameswara monument.In the tropical climatic conditions of eastern India, with high solar radiation, temperature and desiccation, few tolerant organisms can survive.Since L. corticicola has efficient adaptive strategies, it survives well in stone monuments and continues its biodeterioration.
The comparative metabolic profile of L. corticicola and the biological crust of the Parsurameswara monument by LC-ESI-MS/MS shows an interesting result.Presence of dichloroacetic acid, delta-guanidinovaleric acid, lyngbic acid, mycosporine-GABA, porphyra-334 and N-methyl polythene, and the compatible solute trehalose were confirmed by LCMS.Dichloroacetic acid is an organochlorine compound comprising acetic acid carrying two chloro substituents.This is a strongly corrosive organic acid (Meusel and Rehm 1993) and was presented three times more in the biological crust than the cells of L. corticicola.It was previously reported from marine cyanobacteria Asparagopsis taxiformis (Woolard et al. 1979).Dichloroacetic acid has been identified as a metabolite of trichloroacetic acid, and is a toxicologically significant metabolite that can induce liver tumours in B6C3F1 mice (DeAngelo et al. 1991).Algal cells and their metabolites are precursors of dichloroacetic acid biosynthesis (Abd El-Aty et al. 2009).These excreted acids dissolve the mineral matrix by forming salts (acidolysis) or by acting as chelates (complexolysis) (Nowicka-Krawczyk et al. 2022).Our report is the first on dichloroacetic acid in cyanobacterial and biological crust, which might lead to biogeochemical deterioration of the Parsurameswara monument.The presence of trehalose abundance in the crust and cells of L. corticicola indicates how this desiccation-tolerant organism inhabiting the terrestrial environment copes with large and frequent fluctuations in water content in its surroundings.Trehalose is accumulated by anhydrobiotic microorganisms to safeguard biological membranes and proteins from the detrimental consequences of water removal.Trehalose is a non-reducing disaccharide synthesized from (1,4)-linked glucose polymers (De Smet et al. 2000).
Synthesis and degradation of compatible solutes (osmolytes) are used to balance the changing water potential in the cyanobacteria (Pade et al. 2012).Nonreducing sugars might help the cyanobacterial cells cope with dehydration/rehydration cycles (Higo et al. 2007;Kl€ ahn and Hagemann 2011).In the terrestrial cyanobacterium Nostoc commune, the accumulation of trehalose was also observed as a response to desiccation and salt stress (Sakamoto et al. 2009).Though there are many reports on the colonization of cyanobacteria in stone monuments which contribute to the biodeterioration of these ancient structures, there is limited knowledge regarding the control mechanism of cyanobacteria to preserve such monuments (Stephens 2002;Ranalli et al. 2005).The majority of treatments aiming at eliminating phototrophic biofilms typically involve the application of biocides, which can lead to unintended consequences.For instance, these treatments may promote the growth of endoliths (Gaylarde and Morton 1999) or introduce phosphate inputs that facilitate rapid recolonization of cleaned surfaces (Young and Urquhart 1998).A new strategy for monitoring and control of cyanobacterial biofilm on stone monuments using monochromatic light was tried experimentally and gives good results of control/prevention of cyanobacterial biofilm growth in stone monuments (Albertano et al. 2005).In the present study, although we have not worked on the control strategies of cyanobacterial biofilm, we found a dominating cyanobacterium, L. corticicola, and by controlling this cyanobacterium and its microclimate we may be able to develop conservation methods to preserve this ancient stone monument.Since it is a desiccation-tolerant cyanobacterium, having efficient adaptability to desiccation, it can activate trehalose biosynthetic pathways and survive prolonged desiccation.Besides having strongly corrosive organic acids like dichloroacetic acid, it can biocorrode the stone monuments.Hence, by implementing conservation practices that target the microbiome, we have the potential to safeguard this culturally significant stone monument.

Conclusion
Culture-based and metagenomics approaches were used to examine the microbiome of the seventh-century Parsurameswara stone monument.Lyngbya corticicola, an anhydrobiotic cyanobacterium, was identified and demonstrated a remarkable ability to endure six months of desiccation.This particular cyanobacterium produced highly corrosive organic acids, particularly dichloroacetic acid, which can lead to the biocorrosion of stone monuments and their intricate carvings.In addition, we performed a comparative analysis of the metabolic profiles of both the biological crust and cells of L. corticicola.This innovative approach allowed us to investigate the mechanism of biodeterioration in stone monuments.Based on our findings, it can be inferred that the preservation of the ancient Parsurameswara stone monument can be achieved through effective management of L. corticicola.

Figure 2 .
Figure 2. The Krona graph showing the relative abundance microbiome of crust samples of Parsurameswara stone monument (a) at the phylum level (b) cyanobacteria at the genus level.

Figure 3 .
Figure 3. Screening of desiccation-tolerant cyanobacteria isolated from the biological crust of Parsurameswara stone monument.

Figure 4 .
Figure 4. Chlorophyll-a, carotenoid, cellular carbohydrate and cell protein of L. corticicola with response to desiccation for six months.

Table 1 .
Comparative account of metabolites found by LC-MS in the crust samples of Parsurameswara stone monuments and cells of Lyngbya corticicola.