Hamster blood study reveals hidden flaw in gene expression analysis during infection and vaccination

The golden Syrian hamster has long served as the primary experimental model for leptospirosis, the globally distributed zoonotic disease caused by the pathogenic spirochete Leptospira interrogans. Because the species is highly susceptible to acute Leptospira infection, researchers rely on it extensively to dissect host-pathogen interactions and to evaluate candidate vaccines. Yet the model carries a persistent technical weakness: species-specific reagents for protein-based immunological analyses remain scarce, which has pushed reverse transcription quantitative PCR, or RT-qPCR, to the forefront as the principal tool for probing host immune responses at the transcriptional level. Among the available biological matrices, whole blood is especially valuable because it captures systemic immune responses and permits less invasive sampling than terminal tissue collection. A new study now addresses a critical but often overlooked prerequisite for reliable RT-qPCR work in this matrix: the identification of stable internal reference genes.

Reference genes, sometimes called endogenous controls, are the silent anchors of every RT-qPCR experiment. They are assumed to maintain constant basal expression across samples and to remain unaffected by whatever experimental treatment is being investigated. Only when this assumption holds true can the measured expression of target genes, such as inflammatory cytokines, be normalized accurately against them. The trouble, as the study’s authors from the Federal University of Pelotas in Brazil point out, is that many laboratories select commonly used genes such as gapdh and actb on the basis of convention rather than prior validation within the specific biological matrix and experimental context under investigation. Because reference gene stability is tissue- and condition-dependent, rigorous experimental validation is strongly recommended for each model, sample type, and treatment protocol. Building on their previous work that validated reference genes in hamster kidney, liver, and lung tissues, the team turned to blood as a complementary and highly informative matrix for studying systemic host responses during leptospirosis and vaccine evaluation.

The experimental design drew on blood samples from 4-to-6-week-old male and female golden Syrian hamsters randomly assigned to six groups of four animals each. One group was infected with Leptospira interrogans serovar Copenhageni strain Fiocruz L1-130 at approximately 1.5 times ten to the fourth cells per milliliter and monitored until predefined clinical endpoints were reached, including hematuria, nasal bleeding, anorexia, impaired locomotion or breathing, ruffled fur, hunched posture, prostration, or body weight loss exceeding ten percent. The remaining groups received two-dose immunization protocols with ovalbumin adsorbed to two percent aluminum hydroxide, attenuated Salmonella enterica serovar Typhimurium at two times ten to the seventh colony-forming units, or inactivated L. interrogans bacterin at ten to the eighth colony-forming units at fourteen-day intervals. A separate group received two doses of Mycobacterium bovis BCG Pasteur at ten to the sixth colony-forming units at a twenty-one-day interval, while a mock-immunized control group received phosphate-buffered saline following the same schedule as the vaccinated groups. These strategies were deliberately chosen to encompass diverse vaccine platforms, adjuvants, and delivery systems relevant to experimental vaccine development, including approaches used in studies involving Leptospira species, SARS-CoV-2, infectious bronchitis virus, and respiratory syncytial virus.

Whole blood was aseptically collected by cardiac puncture, immediately mixed with RNAlater stabilization solution, and stored at minus eighty degrees Celsius until processing. Total RNA was extracted using a modified protocol combining TRIzol reagent and the RNeasy Mini Kit, with concentrated blood pellets resuspended, lysed, and subjected to chloroform phase separation by centrifugation at twelve thousand times g for fifteen minutes at four degrees Celsius. The aqueous phase was recovered, mixed with absolute ethanol, and applied directly to RNeasy spin columns, after which RNA was eluted twice with thirty-five microliters of pre-warmed RNase-free water at sixty degrees Celsius to maximize yield. For cDNA synthesis, one thousand nanograms of total RNA with A260/280 ratios ranging from 1.92 to 2.08 was reverse-transcribed using random primers in a final volume of forty microliters. Real-time PCR amplification was performed on a QuantStudio 3 system in duplicate reactions containing one hundred nanograms of cDNA, SYBR Green Master Mix, and gene-specific primers, with melting curve analysis from sixty to ninety-five degrees Celsius verifying product specificity.

The panel of eleven candidate reference genes spanned three categories. Four novel candidates had been identified by RNA-seq analysis in hamsters: rps16, aprt, gpx1, and fkbp1a. Four non-conventional reference genes previously reported as stable in different tissues and animal models were included: ppia, ppib, pgk1, and rps29. Three classical reference genes commonly used for RT-qPCR normalization completed the panel: ywhaz, actb, and gapdh. Primers were designed according to MIQE guidelines, and amplification specificity was confirmed by a single band on agarose gel electrophoresis and a single peak in melting curve analysis. Stability was assessed using four widely used algorithms implemented through the RefSeeker R package: BestKeeper, which calculates the mean absolute deviation of raw Ct values; geNorm, which estimates the average pairwise variation or M value; NormFinder, which evaluates intra- and inter-group expression variance; and the comparative delta Ct method, which determines the standard deviation of delta Ct values across gene pairs. A final consensus ranking was calculated using the RefFinder approach, which integrates the geometric mean of the ranks assigned by all four algorithms, with lower geometric mean values indicating greater overall stability.

Across all samples and all six experimental conditions, ppia emerged as the most stable reference gene, ranked first by all algorithms either alone or among the top candidates. gapdh and ppib followed in second and third place in the integrated analysis, respectively. In striking contrast, actb was consistently classified as the least stable candidate across every analytical approach evaluated. The strong performance of ppia and ppib echoes the team’s previous findings in hamster solid tissues, where cyclophilin-encoding genes also ranked among the most stable candidates and outperformed actb in both solid organs and blood. Similar patterns have been reported elsewhere: ppia showed high stability in blood from pigs following Matrix-M administration and in the spleen of mice infected with Listeria monocytogenes. gapdh, meanwhile, has shown variable performance in other models, proving unsuitable in blood leukocytes from pigs infected with Actinobacillus pleuropneumoniae yet stable in the liver and kidney of turbot infected with Edwardsiella tarda.

The instability of actb in hamster blood likely reflects the biology of the matrix itself. Unlike solid tissues, whole blood is a highly dynamic and heterogeneous compartment where shifting leukocyte proportions, such as variations in the neutrophil-to-lymphocyte ratio, can dramatically skew bulk transcriptomic profiles. Furthermore, actb encodes a major cytoskeletal structural protein that is rapidly mobilized during immune cell activation, transendothelial migration, and the cytoskeletal rearrangements required for effective pathogen clearance. The systemic inflammatory responses induced by vaccination and active L. interrogans infection inevitably trigger significant cytoskeletal remodeling and transcriptional upregulation of actb in circulating leukocytes, rendering it highly unsuitable for normalization in this context. In contrast, the high stability of ppia across all matrices relies on the broad and constitutive expression of cyclophilin A, a protein whose basal role in cellular homeostasis and protein folding remains uninterrupted by acute leukocyte mobilization. The RNA-seq-identified genes in the panel performed poorly overall, which may be explained by their susceptibility to modulation during immune challenges: leukocyte activation triggers reactive oxygen species production that directly impacts the antioxidant enzyme gpx1, while immune-driven activation and proliferation of circulating cells likely destabilize the expression of ribosomal proteins such as rps16 and immunophilins such as fkbp1a.

Condition-specific rankings revealed that the optimal reference gene varies with the type of immunological stimulus applied. rps29 was the most stable gene in the L. interrogans bacterin group, whereas ywhaz and aprt ranked first in the aluminum hydroxide and PBS control groups, respectively. ppia was the top-ranked candidate in the BCG Pasteur, attenuated S. Typhimurium, and active L. interrogans infection groups. These differences likely mirror the distinct immune environments induced by each intervention. Leptospiral bacterins and aluminum hydroxide are predominantly associated with humoral responses, whereas BCG and attenuated S. typhimurium promote stronger Th1-biased cellular responses, with BCG additionally inducing trained innate immunity. Active L. interrogans infection represents a markedly different context, characterized by a dysregulated systemic inflammatory response involving pronounced pro- and anti-inflammatory cytokine expression. Because whole blood reflects both the composition and activation state of circulating immune cells, these distinct responses may differentially influence the expression stability of candidate reference genes. ppib ranked among the most stable candidates in several vaccination groups but its performance decreased in samples from infected animals, while gapdh and fkbp1a displayed condition-dependent fluctuations that argue against their use as universal reference genes in hamster blood.

The practical consequences of reference gene choice were demonstrated directly. The team analyzed the expression of tnf-alpha, a key pro-inflammatory cytokine involved in systemic inflammation, in blood samples from hamsters infected with L. interrogans, using the most and least stable genes from the overall ranking as internal controls. Normalization with actb resulted in an approximately tenfold overestimation of tnf-alpha expression compared with normalization using ppia, a statistically significant difference that demonstrates how substantially reference gene selection can distort RT-qPCR data interpretation. Pairwise variation analysis based on geNorm indicated that three reference genes are sufficient for accurate normalization, and no statistically significant difference was observed when tnf-alpha expression was normalized using ppia alone versus the combination of the three most stable genes, ppia, ppib, and gapdh. This integrated analysis provides additional support for using a single reference gene when experimental or practical constraints preclude combining three genes. The authors caution, however, that ppia’s stability may vary according to biological context in other species, as it was reported among the least stable genes in bovine peripheral blood mononuclear cells during haemoparasitic infections, reinforcing that reference gene performance is both species- and condition-dependent.

The study also acknowledges its limitations. The sample size was limited, with four animals per group, and whole blood from vaccinated animals was collected twenty-one days after the booster dose, a time point reflecting a more established systemic phase of the immune response, whereas infected animals were sampled at the predefined clinical endpoint, representing the acute disease stage. The authors note that the consistency of the stability rankings with their previous study and the relatively low variation in Ct values among biological replicates support the robustness of the findings within the experimental conditions evaluated. For vaccinated animals, blood was collected by cardiac puncture at euthanasia rather than through longitudinal sampling, meaning that the systemic perspective captured here reflects terminal collection rather than repeated draws from the same animals. Nevertheless, the systemic view that blood provides, integrating circulating immune cells and soluble mediators, remains particularly relevant when comparing infection and vaccination contexts, as the magnitude and distribution of immune mediators may differ substantially between these conditions.

The broader message for the research community is unambiguous. Reference gene stability in hamster blood varies according to the immunological condition evaluated, reinforcing the need for experimental validation before RT-qPCR normalization in any new model, matrix, or treatment protocol. Among the eleven candidate genes analyzed, ppia showed the highest overall stability across infection and vaccination conditions, supporting its use as a reliable reference gene in comparable experimental settings, and it is recommended specifically for studies involving vectored vaccines based on Salmonella or BCG. In contrast, actb displayed marked instability in most groups, indicating that its conventional use may compromise the accuracy of gene expression analyses in this model. By validating candidate reference genes across diverse vaccine platforms, adjuvants, and delivery systems, this study provides a methodological basis for more accurate and reproducible RT-qPCR analyses in hamster blood and offers a validated foundation for studies of systemic immune responses in the principal experimental model for leptospirosis.

Subject of Research: Validation of stable reference genes for RT-qPCR normalization in blood samples from golden Syrian hamsters under Leptospira infection and vaccination conditions

Article Title: Identification and validation of stable reference genes for RT-qPCR analysis in blood samples from golden Syrian hamsters under different immunological conditions

Article References: Nogueira, J. D., Bunde, T. T., Maia, M. A. C., de Oliveira, N. R., Dellagostin, O. A., & Bohn, T. L. O. (2026). Identification and validation of stable reference genes for RT-qPCR analysis in blood samples from golden Syrian hamsters under different immunological conditions. Molecular Biology Reports, 53(1), Article 1791. https://doi.org/10.1007/s11033-026-12980-8

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12980-8

Keywords: golden Syrian hamster, RT-qPCR, reference genes, ppia, actb, Leptospira interrogans, leptospirosis, tnf-alpha, gene normalization, vaccination, blood transcriptomics, RefSeeker