The hunt for red microba5 March 2020
SM Grace, ML Patchett and GE Norris from the School of Fundamental Sciences at Massey University in New Zealand discuss their search for red ‘microba’ – specifically, the identification of microorganisms involved in ‘red heat’ contamination of salt-cured hides.
‘Red heat’ is an industry term that describes the appearance of brightly coloured patches, streaks or spots on salt-cured products, and is attributed to the occurrence of halophilic microorganisms in the curing salt. Red heat contamination is associated with damage and spoilage of cured hides and skins, understood to be the result of degradative, hydrolytic enzymes secreted by such organisms. This often results in defective leather products, causing wastage and economic loss to industry participants.
Much of the historical work on the microorganisms of red heat has necessitated careful and laborious cultivation, limited to phenotypic analyses. While such methods remain the gold standard of microbial diagnostics, characterisation of slow-growing and difficult-to-culture microorganisms remains challenging. Cultivation-independent molecular techniques such as marker gene sequencing and metagenomic analysis can circumvent some of these difficulties. Additionally, such techniques can directly sample a particular environment and begin to describe the in situ population of microorganisms.
For this study, traditional cultivation was complemented by a cultivation-independent method to identify the composition and diversity of microorganism populations. To control animal variation, a single biological sample of cattle hide was treated with two different salt products. This was carried out to compare the effect of these treatments on the composition and diversity of the microorganism population within that hide.
One of the curing salts is a minimally processed, unsterilised product, shown to produce discolouration characteristic of red heat; while the other has been subjected to a heat sterilisation (stoving) step during processing. The cultivation-independent approach is predicted to show greater microbial diversity by uncovering taxa that are difficult to culture under conventional laboratory conditions. Such microorganisms may be key to the onset of red heat contamination in salt-cured hides and skins. To this end, both approaches primarily employ marker gene sequencing: the phylogenetically-informative 16S ribosomal RNA gene is targeted to identify organisms of bacterial and archaeal origin, while the ITS2 region of the eukaryotic ribosomal gene group is targeted for the identification of fungal organisms.
In cultivation, marker gene sequencing represents a single genome per isolate. In contrast, cultivation independent methods attempt to capture the microbial metagenome; that is, all of the microbial genomes of a mixed population. As a relatively cost-effective and less computationally demanding alternative to whole-genome metagenomics, only the phylogenetic marker genes are targeted, as a proxy for the metagenome of a given sample. This is known as metagenomic amplicon sequencing, or ‘metabarcoding’. A metabarcoding approach was planned for this study, using the Illumina MiSeq high-throughput sequencing platform, and this paper describes the progress made thus far with the identification of cultured isolates and the preparation for the metabarcoding experiment.
Once completed, it’s anticipated that these results may provoke further investigation into whole-genome, functional metagenomics of salt-cured products, to better understand the microbial genes that influence contaminations such as red heat. For the leather and tanning industry, this knowledge may be useful by identifying new microbial targets for both prevention and control. On the basis of the methods used in this study, a PCR-based test could be developed to allow rapid screening of curing salts for microorganisms that either influence or cause red heat in salt-cured products, thus preventing the potential for red heat to occur.
Curing salts used, hide treatment and sampling
Two commercially solar-evaporated salt products, graded for agricultural use, were selected for comparison. The salt referred to as ‘unsterilised’ is a raw, coarse-grained product with a typical moisture content of 1.8%. The salt referred to as ‘sterilised’ is a crushed, washed, kiln-dried product with a typical moisture content of 0.05%. Microorganisms were easily cultured from the unsterilised salt and produced discolouration characteristic of red heat when applied to pieces of bovine hide. Such contamination was not replicated when the sterilised salt was used. The same batch of each salt product was used to cure both of the cattle hides used in this study.
Unshaved cattle hide from the official sampling position (OSP) was divided into two equal-sized pieces, in order to compare the effect of the different salt treatments on the microorganism populations within the same biological sample. Each piece had either sterilised or unsterilised salt applied to the flesh/hypodermal surface at a rate of 50% w/w of the hide sample. Each salted piece was sealed separately inside a clear plastic container with salted side facing up/outwards, and left to cure at room temperature, in ambient light, on the laboratory bench. Hide from two different cattle animals were salt-cured independently in this manner; one hide was used for the cultivation of microorganisms, which was sampled at 90 days of cure. The other hide was used for the metabarcoding experiment, with samples of between 3–5g cut from the hide piece prior to being treated with salt (day zero), then at 24 hours after application of curing salt (day one), then at 10, 20, 40, 50 and 60 days post-salt application. Because of the potentially huge disparity in microbial populations between the hides of different animals, no biological replicates were used. However, to account for differences in extraction and processing, each sampling was done on three different areas of each hide at each time point.
Three different media were used: malt for enrichment of fungal organisms, modified seghal and gibbons for enrichment of fastidious organisms, and lysogeny broth for enrichment of mesophilic bacteria. Each of these media types was prepared with three different concentrations of salt, by diluting a concentrated salt water SW30 stock solution with sterilised, ultrapure water to produce media with a final sodium chloride content of either 20%, 8% or 0.5% (w/v). These amounts were selected as salt concentration optima for enrichment of (extremely) halophilic, moderately halophilic and halotolerant, and non-halophilic microorganisms respectively, resulting in nine different formulations altogether. Solid media was supplemented with 1.5% (w/v) bacteriological agar. All plates were sealed with paraffin film and incubated under a fluorescent bulb, while all liquid cultures were incubated in a table-top shaker in ambient light. Uninoculated controls were incubated to check for presence of environmental contaminants. A modified salt-rice-broth formulation was also used where SW30 stock solution was diluted to a final sodium chloride content of 12% (w/v) with sterilised ultrapure water, to which 10g/L tryptic soy broth and 5g/L acid-hydrolysed casein was added. Two parts of this solution was combined with one part of uncooked, short-grain white rice in a glass tube, then autoclaved to produce a solid, white-coloured growth medium that filled most of the tube.
Colonies selected from enrichment media were transferred to MSG media containing a similar sodium chloride component of either 16%, 8% or 0.5% (w/v) for isolation by streak plate technique. The sodium chloride content was reduced from 20% to 16% (w/v) to ease the preparation of solid media for cultivation of halophilic microorganisms. Aliquots of liquid culture from discrete colonies were frozen in glycerol solution with a final concentration of 15% (w/v).
Samples of approximately 1.5cm2 were cut and pressed onto the surface of solid media. The sample was then halved, with each piece immediately transferred to a flask of sterile brine to wash out microorganisms from within the hide tissue. One flask was prepared with undiluted SW-30 solution (pH 9.0) with a final sodium chloride content of 24% (w/v) and resultant pH of 9.0, to select for extremely halophilic and halotolerant microorganisms. The other flask was prepared by diluting SW-30 solution to achieve a final sodium chloride concentration of approximately 9.6% (w/v), with resultant pH of 8.0, to promote cultivation of halotolerant and slightly halophilic microorganisms. Flasks were incubated for three days with shaking and 120μL of this liquid spread onto the surface of solid media.
To freshly prepared rice-broth tubes, 2.5g of salt sample was added, followed by 3ml of sterilised water, to give an approximate sodium chloride concentration of 14.5% (w/v). Tubes, including uninoculated controls, were loosely capped and incubated. After 10 weeks, 25 sterile loops were used to streak samples onto solid media for isolation.
Approximately 3–5g of hide sample was washed in filter-sterilised ammonium bicarbonate solution (pH 8.0) on rotating arms. To remove particulate matter, the liquid mixture was passed through a nylon membrane with 80μm pore size, into a sterile collection tube. The membrane was washed with a further 10.0ml of the ammonium bicarbonate solution, with collected liquid lyophilised to powder. A ‘reagents-only’ (no hide sample) extraction was then performed at each hide collection time point, as a control for environmental contaminants.
Then 27 brines of 25% (w/v) were made from 100ml of sterilised, ultrapure water and 25g of salt sample, and incubated with gentle shaking for 20–30 minutes. Brines were passed through 10μm pore size nylon membranes to collect cells. A ‘reagents-only’ collection was performed as a control for environmental contamination.
Cultivations after curing
Both of the hide pieces cured with unsterilised salt developed discolouration characteristic of red heat, which first appeared as small, pale pink spots <1.0mm diameter on the flesh surface, along with a faint pink colouration of the salt grains. This appeared at day 23 of cure on the hide used for microorganism cultivation. By day 46 of cure, the colour had deepened to a vivid, fuchsia-like pink and spots were distributed across the entire flesh surface. At day 90 of cure, the pink colouration had deepened and appeared quite dry. Pink spots first appeared on the 40th day of cure on the hide piece sampled for metabarcoding. By day 60 of cure, about a quarter of the flesh surface displayed bright pink-coloured spots and small patches. Furthermore, most of the salt grains displayed a faint pink colouration. Interestingly, after being left for a total of 270 days, patches developed over the majority of the flesh surface and most of the colour had changed from bright pink to a vivid pink-orange, with a glistening, semitransparent appearance. Both of the discoloured pieces had a slight smell of rotting meat and urine; however, it was not overpowering, nor reminiscent of ammonia.
Neither piece showed obvious liquefaction, though the pink-orange coloured piece had a slightly slippery, slimy feel when handling. Some hair was able to be pulled from both hides using tweezers. The two hide pieces cured with sterilised salt did not develop any discolouration. Both had a very faint smell of rotten meat and urine. Neither piece showed any sign of liquefaction, appearing to be dry and relatively inflexible when handled, in comparison to the orangepink contaminated piece. Hair slip was not apparent, but some hair was able to be pulled from the epidermal surface.
A number of colony types were cultivated from the unsterilised-salted hide, with many successfully cultured; however, very few were captured on media containing 0.5% w/v sodium chloride. Coloured colonies were mostly of various shades of pink with some showing pink-orange pigmentation. All were glistening and semi-transparent. Shapes were either round or a ‘fried egg’ appearance; having an opaque, raised centre in an otherwise glistening, semitransparent, irregular margin. These pink colony types occurred exclusively on the 20% and 8% sodium chloride-containing plates. Some opaque, pale yellow colonies were evident on the plates with 8% sodium chloride or less. Many off-white or beige types were isolated, mostly >3mm, with round to irregular margins, occurring most densely on plates containing 8% sodium chloride. From the sterile-salted hide, a few large (20mm) beige-pink, rugose growths with irregular margins were observed, as well as off-white colonies, and many yellow-pigmented colonies, mostly opaque with a glistening appearance. From this hide piece, a total of nine isolates were cultured. Many pale pink, glistening colonies were obtained from the unsterilised salt, along with one plate that developed deep red-pink, glossy, round colonies. However, efforts to culture the red colonies were largely unsuccessful, with only one isolate being cultured. No organisms were isolated from the sterilised salt using the ricebroth enrichment medium.
Additionally, to test the effectiveness of the ricebroth enrichment media, a variety of salt products were screened to detect the presence of coloured organisms. The aspect and onset of discolouration was markedly different between each of these salt samples, with sample 2547-2 producing a bright pink colour within five days of inoculation. The tube inoculated with unsterilised salt product developed colour more slowly, first appearing after 10 days, with others first showing obvious discolouration after 10–21 days of incubation. Most of those that produced discolouration of the enrichment media also produced slow-growing, pigmented, glossy, round to irregular shaped colonies on isolation media. However, non-pigmented colonies were also isolated, which were numerous but much smaller in size compared with the coloured isolates. No isolates were cultured from those tubes that remained white/uncoloured after the incubation period.
DNA band profile
For the six samples taken from the hide pieces prior to being salted (day zero), bacterial marker gene amplification products were clearly detected in the untreated hide. However, amplification of archaeal marker genes across these six samples of untreated hide appeared inconsistent, and fungal marker gene amplification products were largely indistinguishable. Amplification of archaeal and bacterial marker genes was detected in all samples of cured hide taken from between day one and day 60 of cure, with clear differences in the DNA band profiles between the sterilised-salted hide and unsterilised-salted hide.
In samples of hide cured with sterilised salt, the DNA band profile of archaeal marker gene amplification products shows one band approximately 500 base pairs (bp) in length, and another approximately 450bp in length and of similar staining intensity. This appears consistently within the three samples taken at each time point, and between each of the sampling time points. This contrasts with that shown for the unsterilised salted hide samples, where the staining intensity of the 500bp band is greater compared with the faint, almost indistinguishable 450bp band directly below it. The appearance of red heat discolouration on unsterilised salted hide from day 40 of cure did not appear to cause any change to the migration distances of these archaeal marker gene bands in agarose gel electrophoresis. Interestingly, the appearance of red heat discolouration does coincide with a slight increase to the migration distance of the DNA band corresponding to bacterial marker gene amplification products, which is not seen in those obtained from samples of hide cured with the sterilised salt.
In this case, the migration distance of this DNA band remains the same. The 16S rRNA gene is reported to be approximately 1,500bp in length for most bacterial and archaeal organisms. Therefore, marker gene amplification is expected to produce DNA fragments of a mostly uniform length. Speciesspecific size variations in the 16S rRNA gene have been reported, due to accumulation of nucleotide substitutions and deletions, as well as gene truncations and intervening sequences. Thus, variability of amplification product length for this marker gene can be indicative of distinct taxa in these samples.
Results of red heat contamination
We managed to induce red heat contamination in salt-cured cattle hide, and cultivate microorganisms from samples of this hide and also from the salt product used to produce the contamination. We have also cultivated microorganisms from unaffected hide samples. Furthermore, in a cultivation-independent approach, red heat contamination was successfully reproduced in order to directly sample the in situ microorganism population of these hides, in preparation for analysis by high-throughput sequencing of marker gene amplification products. This work has produced a number of interesting results so far. When treated with different salt products, the diversity and composition of both the cultivated and apparent in situ, microorganism population differs markedly within single biological samples of bovine hide. A number of different salt products harboured cultivable microorganisms that were phylogenetically distinct. The marker gene sequences of cultivable microorganisms in the unsterilised salt product are so far quite dissimilar to known reference sequences, suggestive of an as-yet uncharacterised organisms. Analysis of highthroughput sequencing data is expected to clarify this observation.
A change in the composition of the bacterial population coincided with the appearance of red heat contamination, as detected by marker gene amplification, with no such changes apparent in the archaeal and fungal populations. We expect this to be explained by analysis of high-throughput sequencing data generated by these amplification products. The salt product that was treated by heat-sterilisation did not yield cultivable organisms in this study. While microbial phylogenetic marker genes were detected, the viability of these organisms is not clear, indicating the importance of such treatments for the prevention of red heat.
This is an edited extract; the full paper is available on request.