Using FTIR to Detect and Differentiate Microbial Contamination at the Species Level during the Production of Microbial Products

Summary: FTIR-Based Detection and Differentiation of Microbial Contamination during Production

Importance of the Topic

Rapid, reliable detection of microbial contamination is essential in manufacturing processes that use living microorganisms (e.g., industrial fermentations, starter cultures, bioproducts). Contamination can compromise product quality, yield, and biosafety; detecting contaminants at the species level — especially when contaminants belong to the same genus as the production strain — is particularly challenging but critical for targeted corrective actions. Vibrational spectroscopy such as Fourier transform infrared (FTIR) offers a culture-independent, molecularly informative approach that can support high-throughput monitoring and process control.

Objectives and Overview of the Study

The study explored whether diffuse reflectance FTIR (DRS-FTIR) combined with multivariate analysis can detect and distinguish low-level contamination by two Staphylococcus species (S. aureus and S. epidermidis) in cultures of S. cohnii. Key goals were to (1) establish sample preparation and measurement workflows compatible with multi-well, multi-sample throughput, (2) determine practical detection limits for species-level discrimination within the same genus, and (3) validate FTIR findings against quantitative real-time PCR (qPCR) measurements of bacterial composition after co-culture.

Methodology

The experimental workflow comprised controlled co-cultures, targeted sample processing, FTIR-DRS measurement, spectral preprocessing, and multivariate analysis. Critical steps and parameters were:

• Co-culture design: S. cohnii (production strain) was spiked with S. aureus or S. epidermidis at starting proportions ranging from 1.0×10^-2 % to 1.0×10^-7 %, incubated aerobically at 37 °C for 24 h.
• Sample preparation: centrifugation to pellet cells, saline wash, resuspension in ultrapure water, adjustment to 70% ethanol, mechanical homogenization with zirconia beads, deposition of 5 µL aliquots onto a 64-well microfocus plate and drying. Three replicate spots were prepared per sample.
• FTIR measurement mode: diffuse reflectance spectroscopy (DRS) for multi-sample capability and simple cleanup.
• Spectral preprocessing: second derivative (Savitzky–Golay), selection of the fingerprint region (1300–800 cm^-1), and unit vector normalization.
• Data analysis: principal component analysis (PCA) and hierarchical cluster analysis (average linkage, Euclidean distance) on derivative spectra.

Used Instrumentation

Equipment and acquisition settings summarized from the study:

• FTIR system: IRXross coupled with a DRS-8000 diffuse reflectance accessory (custom model).
• Detector: DLATGS (deuterated L-alanine-doped triglycine sulfate) element.
• Spectral resolution: 4 cm^-1; number of averaged scans: 20; apodization: Square Triangle; instrument purged with nitrogen to minimize water and CO2 interference.
• Sample holder: 64-well microfocus plate enabling rapid, multi-spot analysis (measurement time reported ~30 s per sample).

Main Results and Discussion

Key analytical findings from FTIR and supporting qPCR:

• Spectral information: Raw FTIR spectra across samples looked similar, but second-derivative spectra in the fingerprint range (1300–800 cm^-1) revealed reproducible differences attributable to species-specific biomolecular composition (lipids, proteins, polysaccharides).
• Species discrimination: Multivariate analyses resolved four distinct groups in PCA and dendrogram clustering. FTIR reliably detected S. aureus contamination when its starting proportion exceeded approximately 1×10^-5 % (i.e., contamination detectable after 24 h growth when the contaminant had sufficiently increased in relative abundance).
• Limitations with S. epidermidis: FTIR approaches failed to detect S. epidermidis contamination under the tested conditions. qPCR showed that S. epidermidis exhibited poor outgrowth in co-culture with S. cohnii, explaining the inability to detect it by FTIR because its final abundance remained low.
• Growth dynamics: qPCR quantification showed that S. aureus outcompeted S. cohnii and achieved higher relative abundance at end of co-culture when initial S. aureus proportions were above ~1×10^-6 %; this biological amplification was a major factor enabling FTIR detection.
• Throughput and sensitivity trade-offs: The DRS + microfocus plate workflow enabled rapid measurements (~30 s per sample) suitable for monitoring, but effective species-level discrimination depended on both analytical resolution and biological growth behavior during co-culture.

Benefits and Practical Applications

The approach demonstrates a rapid, non-destructive screening option for species-level contamination monitoring in microbial production lines with the following advantages:

• Speed: short per-sample analysis times and compatibility with multi-well plates allow high throughput screening of many samples.
• Molecular specificity: fingerprint-region derivative spectra capture composite biomolecular differences that can separate closely related species when present at sufficient abundance.
• Ease of sample handling: DRS reduces sample preparation complexity and simplifies instrument cleanup compared with transmission/ATR modes for high-throughput workflows.
• Complementary use: FTIR is most effective when integrated with molecular quantification (e.g., qPCR) or culture-based follow-up to confirm low-abundance contaminants and to interpret negative results caused by lack of contaminant outgrowth.

Future Trends and Potential Uses

Opportunities to enhance and extend this methodology include:

• Improved chemometrics: applying supervised classification (PLS-DA, support vector machines, deep learning) and spectral feature selection could improve sensitivity and robustness of species discrimination at lower abundances.
• Targeted sample enrichment: coupling brief selective enrichment protocols or rapid viability assays prior to FTIR could amplify minor contaminants to detectable levels without long culture times.
• Integration into process analytics: embedding FTIR-DRS sampling into in-line or at-line monitoring frameworks can provide near real-time alerts during critical production stages (pre-culture, expansion, harvesting).
• Expanded spectral databases: building comprehensive spectral libraries for production and contaminant strains under standardized culture conditions would aid automated identification across facilities.
• Instrumentation advances: improved detectors, higher-throughput sampling accessories, and automated plate handling will reduce measurement variability and raise throughput further.

Conclusion

Diffuse reflectance FTIR combined with derivative spectral preprocessing and multivariate analysis enabled rapid detection and differentiation of S. aureus contamination in S. cohnii cultures down to a practical threshold (~1×10^-5 % starting contamination), provided the contaminant can grow sufficiently during co-culture. The method was not successful for S. epidermidis under the tested conditions due to its limited growth relative to S. cohnii. FTIR-DRS offers a valuable high-throughput screening tool for microbial production QC when used together with confirmatory molecular assays and when considerations of biological growth dynamics are incorporated into monitoring strategies.

References

1. Lasch P, Naumann D. Infrared Spectroscopy in Microbiology. Encyclopedia of Analytical Chemistry. 2015;3:1–32.