Virus Neutralization and Hemagglutination Inhibition Testing: Principles, Protocols, and Comparative Diagnostics
1. Introduction to Serological Assays for Viral Antibodies
Serological testing forms a cornerstone of veterinary virology, enabling the detection of humoral immune responses following natural infection or vaccination. Among the most widely deployed functional assays are the virus neutralization (VN) test and the hemagglutination inhibition (HI) test. Both assays measure the ability of antibodies to interfere with specific viral functions: VN assesses the capacity of antibodies to block viral infectivity at the cellular level, while HI quantifies antibodies that prevent viral hemagglutinins from binding to erythrocyte receptors [1, 2]. These methods are indispensable for serosurveillance, vaccine efficacy evaluation, and confirmation of infection in the absence of direct pathogen detection [3, 4].
The fundamental distinction between these assays lies in their biological endpoints. VN tests, including the plaque reduction neutralization test (PRNT) and microneutralization (MN) assays, require live virus and permissive cell cultures. They measure the highest serum dilution that reduces viral infectivity by a defined threshold, typically 50% or 90% [4, 5]. HI tests, by contrast, rely on the ability of certain viral surface glycoproteins (hemagglutinins) to agglutinate red blood cells (RBCs). Antibodies specific to these hemagglutinins block agglutination, and the HI titer is the highest serum dilution that completely inhibits hemagglutination [2, 3].
Both methodologies are subject to interference from nonspecific serum inhibitors, necessitating standardized pretreatment protocols. The choice between VN and HI depends on the virus in question, laboratory resources, throughput requirements, and the specific diagnostic or research objective [6, 7].
2. Hemagglutination Inhibition Test: Mechanisms and Methodology
2.1 Biophysical Basis of Hemagglutination
Many enveloped viruses, particularly members of the Orthomyxoviridae (influenza viruses), Paramyxoviridae (Newcastle disease virus, NDV), and Flaviviridae (Japanese encephalitis virus, JEV; duck Tembusu virus, DTMUV), express surface glycoproteins that bind to sialic acid receptors on the surface of erythrocytes [3, 8, 23]. This binding cross-links RBCs, forming a lattice that settles as a diffuse sheet rather than a compact pellet. The hemagglutination (HA) assay quantifies this activity by titrating virus against a standardized RBC suspension. The HI test then uses a fixed concentration of virus (typically 4 hemagglutinating units, HAU) and serial dilutions of serum to determine the antibody titer that prevents lattice formation [2, 32].
2.2 Serum Pretreatment and Removal of Nonspecific Inhibitors
A critical preanalytical step in HI testing is the removal of nonspecific inhibitors of hemagglutination present in serum. These inhibitors are often heat-stable glycoproteins or lipoproteins that can bind viral hemagglutinins or RBCs, producing false-positive inhibition [9, 10]. Standard methods include:
- Receptor-destroying enzyme (RDE) treatment: Derived from Vibrio cholerae neuraminidase, RDE cleaves sialic acid residues from serum glycoproteins, eliminating inhibitors that compete with viral hemagglutinins [5, 10]. This treatment is considered the gold standard for influenza serology [2, 5].
- Kaolin adsorption: Kaolin (aluminum silicate) adsorbs nonspecific inhibitors and some immunoglobulins. The pH of the kaolin-serum mixture can be modified to improve correlation with neutralization tests [40, 41]. However, kaolin may not effectively remove inhibitors from all species, such as equine serum for equine H7N7 influenza virus [9].
- Acetone extraction: Used historically for arbovirus serology, acetone precipitation removes lipids and some inhibitors while preserving specific antibodies [11].
- Heat inactivation: Heating serum at 56 degrees Celsius for 30 minutes inactivates complement and some heat-labile inhibitors, though it is insufficient alone for many viruses [10].
The choice of pretreatment must be validated for each host species and virus combination. For example, RDE treatment is essential for human and avian influenza HI tests but may not be required for all flavivirus HI protocols [2, 36].
2.3 Erythrocyte Selection and Standardization
The species of origin and preparation of RBCs significantly affect HI assay performance. Turkey, chicken, guinea pig, and human type O erythrocytes are commonly used, with species-specific differences in sialic acid receptor expression [28, 39]. Glutaraldehyde-fixed turkey RBCs have been validated for influenza HA and HI assays, offering the advantage of long-term storage at -80 degrees Celsius without significant loss of sensitivity or specificity [39]. Trypsin-modified human erythrocytes have been shown to increase sensitivity in rubella HI tests compared to chicken RBCs [28].
2.4 Assay Protocol and Interpretation
The standard HI protocol involves:
- Preparation of a standardized viral antigen suspension containing 4 HAU in 25 or 50 microliters.
- Two-fold serial dilution of pretreated serum in a 96-well V-bottom plate.
- Addition of the virus suspension to each well, followed by incubation (typically 30-60 minutes at room temperature or 4 degrees Celsius).
- Addition of a standardized RBC suspension (0.5% to 1% v/v).
- Incubation for 30-60 minutes at the optimal temperature for hemagglutination (often 4 degrees Celsius for influenza viruses).
- Reading the endpoint as the highest serum dilution showing complete inhibition of agglutination (a compact RBC button) [2, 3, 32].
The HI titer is expressed as the reciprocal of the highest dilution that completely inhibits hemagglutination. A four-fold rise in titer between paired acute and convalescent sera is considered diagnostic of recent infection [2, 12]. For seroprevalence studies, a single titer cutoff is established using receiver operating characteristic (ROC) curve analysis against a reference standard such as VN [6].
3. Virus Neutralization Tests: Principles and Variants
3.1 Biological Basis of Neutralization
Virus neutralization tests measure the functional capacity of antibodies to block viral entry into permissive host cells. Neutralizing antibodies typically target viral attachment proteins (e.g., hemagglutinin for influenza, E protein for flaviviruses) and prevent receptor binding, membrane fusion, or uncoating [1, 4]. The endpoint is the reduction of a measurable viral effect, such as cytopathic effect (CPE), plaque formation, or viral antigen production.
3.2 Plaque Reduction Neutralization Test (PRNT)
The PRNT is considered the gold standard for many viruses due to its high specificity and quantitative nature [4, 11]. The protocol involves:
- Incubation of serial serum dilutions with a fixed dose of live virus (typically 50-100 plaque-forming units, PFU).
- Addition of the serum-virus mixture to confluent monolayers of permissive cells (e.g., Vero cells for flaviviruses, MDCK cells for influenza).
- Adsorption period, followed by overlay with a semisolid medium (e.g., agarose or carboxymethylcellulose) to restrict viral spread.
- Incubation for a defined period (2-7 days depending on virus).
- Fixation and staining to visualize plaques (foci of infected cells).
- Calculation of the PRNT50 or PRNT90 titer, defined as the serum dilution that reduces plaque count by 50% or 90% relative to the virus control [4, 11].
The PRNT is labor-intensive, requires live virus (posing biocontainment challenges for high-consequence pathogens), and has a turnaround time of several days [4]. Despite these limitations, it remains the reference method for flavivirus serology due to its ability to discriminate between closely related flaviviruses [4, 7].
3.3 Microneutralization (MN) Assays
Microneutralization assays are miniaturized, higher-throughput adaptations of the PRNT, typically performed in 96-well plates. Endpoints are measured by:
- Cytopathic effect (CPE) observation: Visual or spectrophotometric assessment of cell death [5].
- Enzyme-linked immunosorbent assay (ELISA) detection: Quantification of viral antigen (e.g., influenza nucleoprotein) in fixed cells using monoclonal antibodies [5, 24].
- Focus reduction assay: Immunostaining of viral foci using labeled antibodies, counted manually or by automated imaging [1].
MN assays correlate well with HI tests for influenza, though they often demonstrate higher sensitivity, particularly for detecting seroconversion after vaccination or infection [5, 24]. For influenza A(H1N1)pdm09, MN seroconversion rates were 1.5 times higher than HI rates in PCR-positive patients [5]. The MN assay also avoids some of the nonspecific inhibitor issues that plague HI, though RDE treatment of serum is still recommended for optimal sensitivity [5].
3.4 Comparative Performance: VN versus HI
Multiple studies have compared VN and HI across different virus families. Key findings include:
- Influenza viruses: HI and MN show good correlation, but MN is more sensitive for detecting low-level antibodies and seroconversion [5, 24]. The HI test using RDE-treated serum is the method of choice for measuring vaccine-induced antibody levels, with a titer of 1:40 generally considered protective [5, 44].
- Japanese encephalitis virus (JEV): HI and PRNT show significant correlation (R2 = 0.9321) in avian serum samples, with Bland-Altman analysis indicating favorable concordance [4]. HI may serve as a viable substitute for PRNT in large-scale screening [4].
- Duck Tembusu virus (DTMUV): HI and serum neutralization (SN) tests show 100% sensitivity and strong correlation, with HI detecting antibodies as early as day 4 post-infection [7, 8].
- Influenza D virus (IDV) in cattle: HI assays show good correlation with VN (Spearman rank correlation 0.76-0.81), with ROC-derived cutoffs yielding sensitivity of 68-73% and specificity of 94-96% relative to VN [6].
4. Applications in Veterinary Diagnostics
4.1 Avian Influenza and Newcastle Disease
HI testing is the standard serological method for avian influenza virus (AIV) subtype identification and Newcastle disease virus (NDV) surveillance in poultry [3, 23, 30]. The test uses subtype-specific hemagglutinin antigens (e.g., H5, H7, H9) to detect and differentiate antibody responses [3, 13]. For AIV, HI is used for:
- Subtype-specific serosurveillance in commercial and backyard flocks [3].
- Confirmation of infection in conjunction with molecular detection (e.g., RT-PCR).
- Evaluation of vaccine immunogenicity in vaccinated birds [13].
The HI test for NDV was adapted for ostriches using specific RBC and antigen concentrations, demonstrating the need for species-specific optimization [14]. The quantity of virus used in the HI test (HA units) directly affects titer readings, with standardized protocols recommending 4 HAU [32, 33].
4.2 Flavivirus Serology: JEV, DTMUV, and West Nile Virus
Flaviviruses such as JEV, DTMUV, and West Nile virus (WNV) produce hemagglutinins that agglutinate goose or chicken RBCs at specific pH ranges (typically pH 6.0-6.5) [4, 8, 36]. HI tests for these viruses require careful pH optimization and removal of nonspecific inhibitors using acetone extraction or kaolin adsorption [11, 36]. For JEV in avian species, HI shows excellent correlation with PRNT and is recommended for large-scale epidemiological studies [4]. For DTMUV, HI using mouse brain-derived antigen is highly specific, showing no cross-reactivity with AIV, NDV, or other duck viruses [8].
4.3 Coronaviruses and Other Viral Families
HI tests have been developed for bovine coronavirus (BCoV) using mouse brain or cell culture-derived antigens [38]. The test is used for serological diagnosis of coronavirus infection in cattle, though it is less commonly employed than ELISA or VN. HI has also been applied to the diagnosis of Kilham rat parvovirus in laboratory rodents, using specific antigen preparations [34].
4.4 Non-Viral Applications
The HI principle has been adapted for non-viral analytes, including pregnancy diagnosis in non-human primates via detection of urinary chorionic gonadotropin (mCG) [15, 16, 26]. In these assays, hormone-specific antibodies inhibit agglutination of hormone-coated RBCs, providing a rapid, low-cost diagnostic [15].
5. Methodological Considerations and Limitations
5.1 Nonspecific Inhibition and Prozone Effects
Nonspecific inhibitors remain the most significant source of false-positive results in HI testing. The efficacy of RDE, kaolin, or acetone treatment varies by host species and virus [9, 40]. For equine serum, kaolin treatment failed to remove inhibitors for equine H7N7 influenza virus, necessitating alternative approaches [9]. The prozone effect (excess antibody preventing lattice formation) can cause false-negative results at low serum dilutions, particularly in high-titer samples [2].
5.2 Cross-Reactivity and Specificity
HI tests are generally subtype-specific for influenza viruses but may show cross-reactivity within a subtype or between closely related flaviviruses [2, 4]. For flaviviruses, PRNT is preferred for discriminating between infections with different serocomplex members (e.g., JEV versus WNV) [4]. For DTMUV, HI showed no cross-reactivity with other flaviviruses, supporting its use in regions where multiple flaviviruses circulate [7, 8].
5.3 Throughput and Automation
Traditional HI and PRNT are manual, low-throughput methods. Recent advances include:
- High-throughput HI assays: Using recombinant virus-like particles (VLPs) and automated image readers for objective endpoint determination [42, 47]. These assays show excellent precision, linearity, and correlation with MN [42].
- Focus reduction assays: Using immunostaining and automated counting to replace manual plaque counting, increasing throughput for neutralization testing [1].
- Capillary blood sampling: Dried blood spots or capillary blood can be used for HI testing, facilitating field collection and reducing venipuncture requirements [45].
5.4 Biosafety and Reagent Stability
VN tests require live virus, which poses biosafety risks for high-consequence pathogens (e.g., highly pathogenic avian influenza H5N1). HI tests use inactivated or purified hemagglutinin antigens, making them safer for routine use [3, 4]. Glutaraldehyde-fixed RBCs can be stored long-term, reducing reliance on fresh blood supplies [39].
6. Workflow and Decision Framework
The following Mermaid diagram illustrates a decision framework for selecting between HI and VN assays based on diagnostic objectives and available resources.
flowchart TD
A[Diagnostic Objective] --> B{Primary Goal?}
B --> C[Serosurveillance / Vaccine Monitoring]
B --> D[Confirmatory Diagnosis / Differentiation]
C --> E{Throughput Required?}
E --> F[High Throughput]
E --> G[Low Throughput]
F --> H[HI Assay]
G --> I[Microneutralization / PRNT]
D --> J{Virus Family?}
J --> K[Orthomyxoviridae / Paramyxoviridae]
J --> L[Flaviviridae]
K --> M[HI with RDE treatment]
L --> N{Cross-reactivity risk?}
N --> O["Low: HI acceptable"]
N --> P["High: PRNT required"]
H --> Q[Interpret with ROC-derived cutoff]
I --> R[Calculate PRNT50/90 or MN titer]
M --> Q
O --> Q
P --> R
7. Data Interpretation and Statistical Considerations
7.1 Cutoff Determination
For HI tests, the seropositivity cutoff is formally established using ROC curve analysis against a reference standard (typically VN) [6]. For influenza D virus in cattle, a cutoff titer of 10 was determined, yielding sensitivity of 68-73% and specificity of 94-96% [6]. For influenza in humans, a HI titer of 1:40 is widely accepted as a correlate of protection [5, 44]. For veterinary species, protective titers must be established empirically for each virus-host system.
7.2 Correlation and Agreement
Spearman rank correlation is commonly used to assess the relationship between HI and VN titers. Reported correlation coefficients include:
- 0.76-0.81 for IDV in cattle [6].
- 0.90 for H9N2 avian influenza using indirect ELISA versus HI [13].
- 0.93 for JEV in avian sera [4].
- 0.61 for influenza H1N1 using microNT-ELISA versus HI [24].
Bland-Altman analysis is recommended to assess agreement between methods, as correlation alone does not measure concordance [4].
7.3 True Prevalence Estimation
Using HI test sensitivity and specificity estimates, true prevalence can be calculated using stochastic or Bayesian approaches. For IDV in Moroccan cattle, the true prevalence was estimated at 48.44-48.73% using two different HI assays [6].
8. Conclusion
Virus neutralization and hemagglutination inhibition tests remain essential tools in veterinary serology. HI offers simplicity, low cost, and high throughput, making it ideal for surveillance and vaccine monitoring. VN provides higher sensitivity and specificity, particularly for discriminating between closely related viruses, but requires live virus and cell culture infrastructure. The choice between these assays must be guided by the specific diagnostic objective, virus characteristics, available resources, and the need for quantitative versus qualitative data. Proper validation, including ROC-based cutoff determination and assessment of nonspecific inhibitor removal, is critical for reliable results.
References
[1] Feng C, Rowe T, Currier M et al. A DirectView focus reduction assay for high-throughput quantification of neutralizing antibodies against influenza A and B viruses. Cell Rep Methods. 2026. URL: https://pubmed.ncbi.nlm.nih.gov/42242220/
[2] Meijer A, Bosman A, van de Kamp EE et al. Measurement of antibodies to avian influenza virus A(H7N7) in humans by hemagglutination inhibition test. J Virol Methods. 2006. URL: https://pubmed.ncbi.nlm.nih.gov/16271401/
[3] Butaeva IB. Evaluation of the effectiveness of the hemagglutination inhibition (hi) test in the diagnosis of avian influenza. The American Journal of Veterinary Sciences And Wildlife Discovery. 2025. URL: https://www.semanticscholar.org/paper/d36dd1e6321f373cc11ea5bb18a90f7cf50831fd
[4] Cui Li, Jianqing Wan, Deli Wang et al. Comparative Analysis of Hemagglutination Inhibition and Plaque Reduction Neutralization Tests for Japanese Encephalitis Virus Antibody Detection. Viruses. 2025. URL: https://www.semanticscholar.org/paper/327b5d8f2ba86ec8e79b4ed8ec0a18098e00f950
[5] Krivitskaya V, Kuznecova EV, Majorova V et al. Microneutralization reaction compared to hemagglutination inhibition assay to evaluate immunogenicity of influenza vaccines and influenza diagnostics. Russian Journal of Infection and Immunity. 2020. URL: https://www.semanticscholar.org/paper/5117bf8eb4853c3f1f5b021d15431ee190bd797a
[6] Saegerman C, Salem E, Ait Lbacha H et al. Formal estimation of the seropositivity cut-off of the hemagglutination inhibition assay in field diagnosis of influenza D virus in cattle and estimation of the associated true prevalence in Morocco. Transboundary and Emerging Diseases. 2020. URL: https://www.semanticscholar.org/paper/4f5b89db86648dc94d81a84fdf7df382b9f567a3
[7] Tunterak W, Ninvilai P, Prakairungnamthip D et al. Evaluation and comparison of hemagglutination inhibition and indirect immunofluorescence tests for the detection of antibodies against duck Tembusu virus. Transboundary and Emerging Diseases. 2022. URL: https://www.semanticscholar.org/paper/3dc0a7088120ec704f91d4a9a808a5b19b428210
[8] Wang X, Yang Z, Wang X et al. Development of a Hemagglutination Inhibition Assay for Duck Tembusu Virus. Avian diseases. 2019. URL: https://www.semanticscholar.org/paper/faec9e3b680ec9887ec2c4facf065dea333d69c3
[9] Boliar S, Stanislawek W, Chambers T. Inability of Kaolin Treatment to Remove Nonspecific Inhibitors from Equine Serum for the Hemagglutination Inhibition Test Against Equine H7N7 Influenza Virus. Journal of Veterinary Diagnostic Investigation. 2006. URL: https://www.semanticscholar.org/paper/f25fd14918e28adf2c784cc529ce8357beebbf87
[10] Jordan WS, Oseasohn R. The use of RDE to improve the sensitivity of the hemagglutination-inhibition test for the serologic diagnosis of influenza. Journal of Immunology. 1954. URL: https://www.semanticscholar.org/paper/929f1e4e3e90a358fea660047ac7296a2d95b935
[11] Chanock R, Sabin A. The hemagglutinin of St. Louis encephalitis virus. III. Properties of normal inhibitors and specific antibody; use of hemagglutination-inhibition for diagnosis of infection. Journal of Immunology. 1953. URL: https://www.semanticscholar.org/paper/608e08a7f10e001cd2781505ff6ab26fdf1d76da
[12] Julkunen I, Pyhälä R, Hovi T. Enzyme immunoassay, complement fixation and hemagglutination inhibition tests in the diagnosis of influenza A and B virus infections. Purified hemagglutinin in subtype-specific diagnosis. Journal of Virological Methods. 1985. URL: https://www.semanticscholar.org/paper/706751e1c532dc9b2eee5806d6471300c0fdcc9e
[13] Madani R, Hezarosi M, Golchinfar F. Unveiling Indirect ELISA Test against Nucleoprotein of H9N2 Comparing With Hemagglutination Inhibition Test. Archives of Razi Institute. 2024. URL: https://www.semanticscholar.org/paper/dbddc1b6af9a95110d98f4c10cfe17a74cf1f6a3
[14] Carrasco AOT, Neto OCF, Lages SLS et al. ADAPTATION OF HEMAGGLUTINATION INHIBITION TECHNIQUE (HI) FOR THE DIAGNOSIS OF NEWCASTLE DISEASE IN OSTRICHES (Struthio camellus). 2008. URL: https://www.semanticscholar.org/paper/2acb6ce7029c8478b2ebedf6956f33e35e6bb5d5
[15] Hodgen G, Ross G. Pregnancy diagnosis by a hemagglutination inhibition test for urinary macaque chorionic gonadotropin (mCG). Journal of Clinical Endocrinology and Metabolism. 1974. URL: https://www.semanticscholar.org/paper/cfcdf643c3e3425c54cd10887bba7a8a8519a8be
[16] Hodgen G, Wolfe LG, Ogden J et al. Diagnosis of pregnancy in marmosets: hemagglutination inhibition test and radioimmunoassay for urinary chorionic gonadotropin. Laboratory animal science. 1976. URL: https://www.semanticscholar.org/paper/7fbbd1f2b53f74fc61af5ade4b506ec3ff0db4f1
[17] Kunz C, Hofmann H. [Early diagnosis of tick-borne encephalitis in the hemagglutination-inhibition-test by treatment of serum with 2 mercaptoethanol]. Zentralblatt fur Bakteriologie, Parasitenkunde, Infektionskrankheiten und Hygiene. Erste Abteilung Originale. Reihe A: Medizinische Mikrobiologie und Parasitologie. 1971. URL: https://www.semanticscholar.org/paper/ad949ebdb0ae58adad3348580d8f4570ebc924b1
[18] Kurskaia OG, Durymanov AG, Sharshov K et al. [Hemagglutination inhibition test for retrospective diagnosis of avian influenza in mammals]. Zhurnal mikrobiologii, epidemiologii, i immunobiologii. 2009. URL: https://www.semanticscholar.org/paper/39608d1fdd5e78b6be16281af217227f1875b2a3
[19] Prasittikhet K, Thawatsupha P, Kijphati R et al. Comparison of Hemagglutination Inhibition Test (HI) and Enzyme Linked Immunosorbent Assay (ELISA) for Diagnosis of Influenza Virus Infection. 2012. URL: https://www.semanticscholar.org/paper/ec2b7a9163da46c836bc7934bf11c69f80720cd0
[20] Lennette EH, Schmidt N, Magoffin R. The hemagglutination inhibition test for rubella: a comparison of its sensitivity to that of neutralization, complement fixation and fluorescent antibody tests for diagnosis of infection and determination of immunity status. Journal of Immunology. 1967. URL: https://www.semanticscholar.org/paper/243c85cd26f7ffd2ce98e02bad88b2652d48770a
[21] Turner R, Lathey J, van Voris LP et al. Serological diagnosis of influenza B virus infection: comparison of an enzyme-linked immunosorbent assay and the hemagglutination inhibition test. Journal of Clinical Microbiology. 1982. URL: https://www.semanticscholar.org/paper/f871bd290914ecead5470b6a265a2a9237666498
[22] Monto AS, Maassab H. Ether treatment of type B influenza virus antigen for the hemagglutination inhibition test. Journal of Clinical Microbiology. 1981. URL
Disclaimer: This article is for educational and informational purposes only. It is not intended to substitute for professional veterinary advice, diagnosis, treatment, or regulatory guidance. Always consult a licensed veterinarian or qualified specialist regarding animal health, disease diagnosis, and therapeutic decisions.