Oxidative Stress Markers: Lipid, Protein and DNA Assays Compared

Oxidative stress markers are among the most widely published and most frequently misread endpoints in redox biology. This guide compares the three damage classes researchers actually measure — lipid peroxidation, protein oxidation and DNA oxidation — explains why a TBARS value is not a malondialdehyde (MDA) value, and covers the sample handling that decides your result before the plate is ever read. Every kit and antibody named below was checked against the live BioHippo ELISA kit catalog on 28 August 2026.

Oxidative Stress Markers vs Live-Cell Probes: A Footprint Is Not a Rate

Probe-based methods report on oxidant flux at the moment you look. Damage markers report on something different: the accumulated chemical modifications that oxidants left behind on lipids, proteins and DNA. The two answer different questions, and confusing them is the most common interpretive error in the field.

A damage marker integrates over time. It survives the oxidant that produced it — often by hours or days — which is exactly why it works in plasma, urine, serum and fixed tissue where live-cell probes cannot go. That same property is its limitation: a marker level reflects the balance between the rate of damage and the rate of repair or clearance, so an unchanged marker does not prove that oxidant production was unchanged.

Three practical consequences follow. Damage markers are the right choice for clinical samples, archived material, in vivo studies and any experiment where the readout must survive collection. They are the wrong choice for kinetics, for subcellular localization, or for anything measured in minutes. And because different oxidants attack different macromolecules, the marker class you select is itself a hypothesis about the chemistry you expect.

If you need real-time flux rather than accumulated damage, start with Part 1 of this series: Reactive Oxygen Species Assay Tools: A Guide to Detecting ROS Across the Redox Workflow. This note goes deep on the damage endpoints that guide only introduced.

Lipid Peroxidation Markers: MDA, TBARS, 4-HNE and 8-Isoprostane

Polyunsaturated fatty acids are the most oxidizable structures in the cell. Hydrogen abstraction from a bis-allylic carbon starts a self-propagating radical chain that converts one initiating event into many oxidized lipids, which is why a lipid peroxidation assay is sensitive — and why it amplifies artifacts just as efficiently.

MDA — malondialdehyde

A three-carbon dialdehyde generated from arachidonic acid, DHA and other fatty acids carrying three or more double bonds. MDA is the most-cited lipid peroxidation product. It is also reactive enough to form adducts with lysine residues and with deoxyguanosine, so a fraction of total MDA in any sample is protein- or DNA-bound rather than free. Immunoassays and chromatographic methods differ in how much of that bound pool they recover, which is one reason absolute MDA values vary widely between publications.

4-HNE — 4-hydroxy-2-nonenal

An α,β-unsaturated aldehyde derived from ω-6 PUFAs such as linoleic and arachidonic acid. 4-HNE is longer-lived and more diffusible than the radicals that produce it, and forms stable Michael adducts with cysteine, histidine and lysine side chains. This is the key interpretive point: most 4-HNE in a biological sample is protein-bound, so immunoassays and antibodies generally detect HNE–protein adducts rather than the free aldehyde. That makes 4-HNE a marker of downstream protein modification as much as of the lipid event that started it, and it is why the same antibody works well for tissue localization by immunohistochemistry.

8-iso-PGF_2α — 8-isoprostane

The most chemically defensible lipid peroxidation marker available in immunoassay format. F₂-isoprostanes are chemically stable, present at quantifiable baseline concentrations in plasma and urine, and formed by a well-characterized non-enzymatic free-radical rearrangement of arachidonic acid. They have been described in the free-radical literature as the “gold standard” biomarker of lipid peroxidation in vivo (Musiek et al., Lipids 2005). Urinary measurement is particularly useful because it sidesteps the ex vivo peroxidation that plagues plasma lipid markers. One caveat belongs in any methods section: F₂-isoprostane isomers can also arise through cyclooxygenase activity, so isoprostane elevation is not exclusively non-enzymatic in tissues with high COX expression. Note also that immunoassay formats measure a single isoprostane and are less specific than GC-MS or LC-MS/MS (Michel et al., Ann Biol Clin 2008).

Lipid peroxidation kits and antibodies verified in the BioHippo catalog

• Human MDA (Malondialdehyde) ELISA Kit — ELK Biotechnology; additional species include mouse, rat, pig, bovine, chicken and zebrafish. $344.50
• QuantiChrom™ TBARS Assay Kit — BioAssay Systems; colorimetric, species-independent. $297.67
• Human TBARS ELISA Kit — ELK Biotechnology. $416.00
• Human HNE (4-Hydroxynonenal) ELISA Kit — ELK Biotechnology; microsample format also available. $416.00
• MOLECULAR SIGNATURE® 4-Hydroxynonenal Antibody — StressMarq Biosciences; for adduct detection by Western blot and IHC. $429.51
• Anti-4R-HNE/4-HNE Recombinant Antibody (R310) — AtaGenix Laboratories. $291.00
• Human 8-iso-PGF_2α (8-isoprostane) ELISA Kit — Fine Test; mouse and rat versions available. $520.00

Why a TBARS Assay Value Is Not an MDA Value

In the TBARS assay, thiobarbituric acid is reacted with the sample under acid at 95–100 °C; MDA forms a pink MDA–TBA₂ adduct read at 532 nm, or fluorometrically at excitation 515 nm / emission 553 nm. The method is cheap, fast and needs no specialized instrumentation, which explains its dominance. The name — thiobarbituric acid reactive substances — is deliberately hedged, and the hedge matters.

Thiobarbituric acid is not selective for MDA. Under the assay’s acidic, high-temperature conditions it also reacts with sugars, bile pigments including bilirubin, amino acids, deoxyribose and a range of other aldehydes, all of which absorb in the same region. Sample matrices rich in these interferents return inflated readings that have nothing to do with lipid peroxidation. The limited analytical specificity of the assay is acknowledged even in method papers that recommend its use (Aguilar Diaz De Leon & Borges, J Vis Exp 2020).

Worse, the heating step is not a passive extraction. Lipid hydroperoxides that were intact in the original sample decompose during incubation and generate MDA that did not exist before the assay began. The assay partly creates the analyte it measures, and the size of that contribution scales with how much oxidizable lipid the sample contains. When plasma is measured both ways, chromatographically resolved MDA is consistently and significantly lower than the TBARS value, and the between-individual variation visible by HPLC disappears in the TBARS readout — which is precisely the signal most studies are trying to detect (Moselhy et al., J Lipid Res 2013; Templar et al., Nephrol Dial Transplant 1999).

What to do: include a chain-breaking antioxidant such as BHT in the reaction to suppress peroxidation during heating; resolve the MDA–TBA adduct chromatographically when specificity matters; and report the endpoint as TBARS, not as MDA, unless it has been chromatographically confirmed. Reviewers increasingly ask for exactly this distinction, and a TBARS-only result supporting a mechanistic claim is a common source of revision requests.

Protein Oxidation Markers: Carbonyls, 3-Nitrotyrosine and AOPP

Protein oxidation markers have one decisive advantage over lipid markers: the principal modifications are irreversible and are not efficiently repaired. They accumulate, which makes them better integrators of chronic oxidative burden and less sensitive to the short-term fluctuations that make lipid markers noisy.

Protein carbonyls — the general-purpose protein carbonyl assay

Carbonyl groups are introduced onto proline, arginine, lysine and threonine side chains by metal-catalyzed oxidation, and by adduction of reactive aldehydes such as 4-HNE, so the endpoint aggregates several damage routes into one number. Detection almost always proceeds through derivatization with 2,4-dinitrophenylhydrazine (DNPH) to form a dinitrophenylhydrazone, quantified either by absorbance around 375 nm or immunologically with an anti-DNP antibody. Because carbonyls are stable and cumulative, they suit aging, chronic disease and long-exposure models.

3-Nitrotyrosine — a reactive nitrogen species marker, not a general ROS marker

Formed when nitric oxide-derived species nitrate tyrosine residues, principally via peroxynitrite generated from the near-diffusion-limited reaction of NO• with superoxide, and via myeloperoxidase acting on nitrite. Read the marker class carefully: 3-nitrotyrosine reports reactive nitrogen species, not ROS generally. That specificity is its value — a nitrotyrosine signal implicates the NO•/superoxide axis rather than generic oxidant load — and it is also why substituting it for a general oxidative damage marker misstates the chemistry.

AOPP — advanced oxidation protein products

AOPP are dityrosine-containing cross-linked protein modifications, largely albumin-derived, formed mainly via the myeloperoxidase reaction and therefore pointing to activated neutrophils and monocytes (Sebeková et al., J Ren Nutr 2012). AOPP consequently behaves as much like an inflammatory readout as an oxidative one, and it has an established literature in chronic kidney disease and dialysis cohorts, where levels run roughly 2-fold above healthy controls. If your model involves phagocyte recruitment, AOPP adds information that carbonyls alone will not.

Choosing between them. These three are not interchangeable. Carbonyls answer “how much cumulative protein oxidation is there?”; 3-nitrotyrosine answers “is nitrosative chemistry involved?”; AOPP answers “is myeloperoxidase-driven, phagocyte-associated oxidation involved?” Running carbonyls plus one mechanism-specific marker is usually more informative than running two general markers.

Protein oxidation kits and antibodies verified in the BioHippo catalog

• Human PC (Protein Carbonyl) ELISA Kit — ELK Biotechnology; rat, rabbit and microsample formats available. $416.00
• Human Protein Carbonyl (PC) ELISA Kit — Bioassay Technology Laboratory; fish version available. $320.00
• 3-NT (3-Nitrotyrosine) ELISA Kit — Elabscience Biotechnology. $154.50
• MOLECULAR SIGNATURE® Nitrotyrosine Antibody — StressMarq Biosciences. $360.50
• U-Blot® Pan-3-Nitrotyrosine Rabbit mAb — UcallM Biosciences; mouse mAb and rabbit pAb clones also listed. $772.50
• Human AOPP (Advanced Oxidation Protein Products) ELISA Kit — ELK Biotechnology; mouse, rat, horse, cattle and chicken available. $416.00

For low-abundance analytes and small sample volumes, compare formats in the high-sensitivity ELISA kit collection, and see the matching antibody catalog for Western blot and IHC work.

DNA Oxidation Markers: 8-OHdG in Tissue DNA and in Urine

Guanine has the lowest oxidation potential of the four DNA bases, so it is preferentially attacked, and its C8-hydroxylation product 8-hydroxy-2’-deoxyguanosine (8-OHdG, also written 8-oxo-dG) has become the field’s default nucleic acid damage marker.

8-OHdG is formed in genomic DNA and in the nucleotide pool, then excised by OGG1-initiated base excision repair and excreted in urine. That routing matters: urinary 8-OHdG reflects whole-body repair activity, while tissue or leukocyte DNA gives you the residual lesion load remaining in that compartment. Both readouts are legitimate; they are simply not the same measurement, and papers frequently conflate them.

Immunoassay values and chromatographic values do not agree

This is a documented, decades-old discrepancy rather than a vendor-specific defect. In matched clinical samples, two variants of a commercial 8-oxo-dG ELISA correlated with LC-MS/MS yet still overestimated concentrations, and neither reproduced the between-group difference that the chromatographic method detected (Garratt et al., Free Radic Biol Med 2010). A later multi-laboratory exercise using a single kit, standardized temperature control and solid-phase extraction improved agreement but did not eliminate it, and identified saccharides including D-glucose and D-galactose among the interfering urinary constituents; the authors concluded that ELISA still should not be treated as a robust alternative to chromatography (Rossner et al., Free Radic Biol Med 2016).

What to do: treat ELISA-derived 8-OHdG as a relative measure for comparing treatment groups run on the same kit, lot and matrix — not as an absolute concentration. Never compare an ELISA value to a published chromatographic value. State the platform explicitly in the methods. Solid-phase extraction of urine before assay, overnight antibody incubation at 4 °C, and normalization to urinary creatinine have been shown to bring ELISA and LC-MS/MS into close agreement, and are worth adopting (Rossner et al., Biochem Biophys Res Commun 2013).

Extraction generates the lesion you are measuring

This applies to genomic DNA workflows only — measuring 8-oxo-dG in DNA isolated from tissue or leukocytes, where the sample must be lysed, the DNA purified, and the DNA hydrolyzed to nucleosides before quantification. It does not apply to urinary 8-OHdG, where the excised nucleoside is measured directly and no extraction step exists.

In that genomic workflow, guanine oxidation continues during isolation. Phenol-based extraction and adventitious transition metals both drive artifactual 8-oxo-dG formation, and the artifact can rival or exceed the endogenous signal. The European Standards Committee on Oxidative DNA Damage (ESCODD) was formed specifically to address this, and its inter-laboratory work found method-dependent medians differing by roughly an order of magnitude between chromatographic and enzymic (FPG/comet) approaches on identical HeLa cell samples (ESCODD, Free Radic Biol Med 2003). A follow-up validation study across eight European countries concluded that the true background level of base oxidation in DNA is orders of magnitude lower than had often been claimed (Gedik & Collins, FASEB J 2005).

What to do: include an iron chelator such as deferoxamine in the lysis and extraction buffers, keep the sample cold and the workup short, and avoid phenol where a column or salting-out method will serve. Extract every comparison group in the same batch with the same protocol — a batch effect here is indistinguishable from a treatment effect.

DNA damage marker kits verified in the BioHippo catalog

• Human 8-OHdG (DNA Damage) ELISA Kit — ELK Biotechnology; rat, mouse and chicken versions available. $344.50
• 8-OHdG High Sensitivity ELISA Kit — Fine Test; mouse and rat high-sensitivity versions also listed. $520.00
• Human 8-Hydroxy-desoxyguanosine (8-OHdG) ELISA Kit — Bioassay Technology Laboratory. $320.00

Pre-Analytical Control Decides Your Oxidative Stress Marker Result

Every marker in this guide is a chemical modification that can continue to form after the sample leaves the animal, the patient or the dish. Oxidation does not pause for your freezer. In practice, differences in collection and storage generate more between-group variance in damage marker studies than the biology under investigation — and unlike biological variance, this kind is systematic and directional.

1. Get a chelator into the first buffer the sample meets. Free iron and copper drive Fenton chemistry in whatever vessel the sample sits in, so this applies to every matrix and not only to blood. Include EDTA in homogenization and lysis buffers for tissue and cells; draw blood into EDTA tubes rather than clot tubes, since clotting also holds the sample at room temperature and releases platelet and leukocyte contents into it.
2. Add a chain-breaking antioxidant to lipid samples. BHT suppresses the propagation reactions that inflate MDA, TBARS and isoprostane readings during storage and during the assay’s own heating step. Add it before freezing, not at assay time.
3. Reject hemolyzed plasma and serum samples. Heme iron is a potent pro-oxidant catalyst and hemoglobin absorbs strongly in the visible range used by colorimetric assays. Hemolysis inflates the reading twice over, chemically and optically. The tissue equivalent is incomplete perfusion: residual blood in a homogenate does the same thing.
4. Store at −80 °C and aliquot on first thaw. −20 °C is not adequate for damage markers over months. Freeze–thaw cycles measurably increase carbonyl and TBARS values, so aliquot to single-use volumes rather than repeatedly sampling one tube.
5. Keep everything cold and fast. Process on ice, minimize the interval between collection and freezing, and hold that interval constant across groups. A consistent 30-minute delay is far better than an inconsistent 5-to-40-minute one.
6. Batch and randomize. Run all comparison groups on the same kit lot, in the same run, with treatment and control interleaved across the plate. Plate position and lot-to-lot calibration shifts are both large enough to manufacture a result.
7. Normalize appropriately, and report the normalizer. Protein carbonyls to total protein; urinary markers to creatinine; tissue markers to wet weight or DNA content. An unnormalized damage marker is uninterpretable across samples.

Matrix-specific additions

The seven steps above are general. What changes between sample types is where the oxidation happens, and therefore where you intervene.

A useful internal control: include a pooled reference sample, aliquoted once and stored alongside your study samples, on every plate across the whole study. Its drift over time is a direct readout of how much of your between-group signal is pre-analytical rather than biological.

Matching the Oxidative Stress Marker to Your Question

When Immunoassay Is Not the Right Format

BioHippo’s oxidative damage coverage is built on immunoassay format, and the species depth behind it is unusual: MDA, TBARS, AOPP and 8-OHdG kits span human, mouse, rat, rabbit, pig, sheep, goat, horse, cattle, chicken, duck, fish and zebrafish across the listed vendors, with microsample and high-sensitivity variants in several lines. For plasma, serum, urine and lysate work — where most damage marker studies live — that breadth is usually the deciding factor, and an unusual species is worth asking about rather than assuming unavailable.

Immunoassay is not the right format for every method described above. Plan for these alongside your kits rather than in place of them:

• Immunohistochemical 8-OHdG localization in tissue requires an anti-8-OHdG lesion antibody. Note that OGG1 antibodies detect the repair glycosylase that excises the lesion, not the lesion itself — the two answer different questions and are easy to conflate when searching.
• Carbonylated protein detection by immunoblot (OxyBlot-style) needs DNPH derivatization with an anti-DNP antibody. Protein carbonyl coverage here is quantitative ELISA, which gives total burden rather than which proteins carry it.
• Per-cell strand break distribution calls for the comet assay, a useful companion to bulk 8-OHdG when cell-to-cell heterogeneity matters.
• Absolute quantification by LC-MS/MS or HPLC-ECD requires authentic MDA, 8-oxo-dG or isoprostane standards for calibration. This is the route when you need absolute concentrations rather than group comparisons.
• Non-ELISA lipid peroxidation is currently served by the QuantiChrom™ TBARS kit; a standalone colorimetric or fluorometric MDA assay is not part of the line.

None of this rules out the kits above — most studies pair a quantitative immunoassay across all samples with a lower-throughput confirmatory method on a subset. If you are unsure which combination your endpoint needs, request a quote and describe the model, and we will size the panel with you.

Frequently Asked Questions

How do you measure oxidative stress in a biological sample?

You choose between two endpoint types. Live-cell probes measure oxidant production in real time but only in living cells. Oxidative stress markers — the damage endpoints in this guide — measure the accumulated chemical modifications left on lipids, proteins or DNA, and work in plasma, serum, urine, tissue and archived material. Most published studies use a damage marker, normalized to protein, creatinine or DNA content.

Is TBARS the same as MDA?

No. TBARS is an operational endpoint — everything in the sample that forms a coloured adduct with thiobarbituric acid under hot acid — and MDA is only one contributor. Sugars, bilirubin, deoxyribose, amino acids and other aldehydes all react. Chromatographic MDA values in the same plasma samples are consistently lower than TBARS values. Report the endpoint as TBARS unless you have confirmed it chromatographically.

Which oxidative stress marker should I choose for a chronic disease model?

Protein carbonyls, because the modification is irreversible and cumulative, so it integrates chronic burden better than the noisier lipid markers. Add one mechanism-specific marker — 3-nitrotyrosine if you suspect nitrosative chemistry, AOPP if phagocyte recruitment is part of the model — rather than a second general marker.

Can I compare my 8-OHdG ELISA result to a published LC-MS/MS value?

No. Immunoassay and chromatographic 8-OHdG values are not interchangeable; ELISA typically returns higher concentrations because the antibody cross-reacts with structurally related urinary constituents. Use ELISA for relative comparison between groups run on the same kit and lot, and state the platform in your methods.

Does an unchanged damage marker mean there was no oxidative stress?

No. A damage marker level is the balance of damage formation against repair and clearance. A flat marker is compatible with unchanged oxidant production, or with increased production matched by increased repair. To separate those, pair the damage marker with a flux measurement or a repair readout.

References

Literature identifiers below were verified in PubMed. Where a claim rests on established methodological consensus rather than a single study, that is stated in the text.

1. Aguilar Diaz De Leon J, Borges CR. Evaluation of oxidative stress in biological samples using the thiobarbituric acid reactive substances assay. J Vis Exp. 2020;(159). doi:10.3791/61122
2. Moselhy HF, Reid RG, Yousef S, Boyle SP. A specific, accurate, and sensitive measure of total plasma malondialdehyde by HPLC. J Lipid Res. 2013;54(3):852–858. doi:10.1194/jlr.D032698
3. Templar J, Kon SP, Milligan TP, Newman DJ, Raftery MJ. Increased plasma malondialdehyde levels in glomerular disease as determined by a fully validated HPLC method. Nephrol Dial Transplant. 1999;14(4):946–951. doi:10.1093/ndt/14.4.946
4. Musiek ES, Yin H, Milne GL, Morrow JD. Recent advances in the biochemistry and clinical relevance of the isoprostane pathway. Lipids. 2005;40(10):987–994. doi:10.1007/s11745-005-1460-7
5. Michel F, Bonnefont-Rousselot D, Mas E, Drai J, Thérond P. Biomarkers of lipid peroxidation: analytical aspects. Ann Biol Clin (Paris). 2008;66(6):605–620. doi:10.1684/abc.2008.0283
6. Sebéková K, Klenovicsová K, Ferenczová J, Hedvig J, Podracká L, Heidland A. Advanced oxidation protein products and advanced glycation end products in children and adolescents with chronic renal insufficiency. J Ren Nutr. 2012;22(1):143–148. doi:10.1053/j.jrn.2011.10.022
7. Garratt LW, Mistry V, Singh R, et al. Interpretation of urinary 8-oxo-7,8-dihydro-2’-deoxyguanosine is adversely affected by methodological inaccuracies when using a commercial ELISA. Free Radic Biol Med. 2010;48(11):1460–1464. doi:10.1016/j.freeradbiomed.2010.02.017
8. Rossner P, Mistry V, Singh R, Sram RJ, Cooke MS. Urinary 8-oxo-7,8-dihydro-2’-deoxyguanosine values determined by a modified ELISA improves agreement with HPLC-MS/MS. Biochem Biophys Res Commun. 2013;440(4):725–730. doi:10.1016/j.bbrc.2013.09.133
9. Rossner P, Orhan H, Koppen G, et al. Urinary 8-oxo-7,8-dihydro-2’-deoxyguanosine analysis by an improved ELISA: an inter-laboratory comparison study. Free Radic Biol Med. 2016;95:169–179. doi:10.1016/j.freeradbiomed.2016.03.016
10. ESCODD (European Standards Committee on Oxidative DNA Damage). Measurement of DNA oxidation in human cells by chromatographic and enzymic methods. Free Radic Biol Med. 2003;34(8):1089–1099. doi:10.1016/s0891-5849(03)00041-8
11. Gedik CM, Collins A. Establishing the background level of base oxidation in human lymphocyte DNA: results of an interlaboratory validation study. FASEB J. 2005;19(1):82–84. doi:10.1096/fj.04-1767fje
12. BioHippo. Reactive Oxygen Species Assay Tools: A Guide to Detecting ROS Across the Redox Workflow (Part 1 of this series).

Product availability, pricing, species reactivity and sensitivity were verified against the live BioHippo catalog on 28 August 2026 and are subject to change. Confirm sample compatibility, detection range and the intended-use statement on each product page before purchase. Methodological guidance summarizes published literature and does not replace validation in your own matrix. Figures 1–3 are illustrative schematics, not experimental data.