iPSC Culture Explained: Principles, Workflow and When to Use It

iPSC culture gives you a renewable supply of human cells with a defined genetic background, in cell types you cannot biopsy — and it costs more, drifts more, and fails more quietly than any other cell culture you will run. The hard part of iPSC culture is not making the cells; it is keeping them the same. This guide covers the principles that govern the pluripotent state, the eight-step workflow, the decision of whether to culture at all, and the applications where the cost is justified.

Unlike an immortalized line, an iPSC culture has no stable resting state you can neglect for a week. Pluripotency is actively held in place by growth factors you add to the medium, and it is actively eroded by density, by mechanical stress, by spontaneous differentiation at colony edges, and by the slow selection of genetic variants that grow faster than the cells you started with. Left alone, the culture does not stay still — it moves toward whatever the conditions reward. That single fact changes how you plan an experiment: a result at passage 18 is not automatically comparable to the same experiment at passage 45, and a phenotype in one donor line is not automatically a disease phenotype.

iPSC culture overview: three phases, three different jobs

An iPSC culture overview is easiest to hold in your head as three phases that run under genuinely different rules. Reagents, timescales, failure signatures and QC all change between them, and treating the whole thing as “cell culture” is the most common reason a project stalls.

Most labs entering the field run only phases 2 and 3 — they obtain an established line rather than reprogramming their own, because deriving a line is a project in itself and the resulting line still has to be characterized from scratch. Some labs run only phase 3, or skip it entirely by buying cells that are already differentiated. All three positions are legitimate, and the section on when to use iPSC culture below treats the choice as the real decision it is.

Five iPSC culture principles that explain most outcomes

These five iPSC culture principles are not a protocol. They are the reasons the protocol looks the way it does — and if you understand them, most troubleshooting becomes obvious rather than mysterious.

Principle 1 — Pluripotency is a maintained state, not a stable one

Human iPSCs sit in a “primed” pluripotent state that depends on continuous FGF2 and TGF-β/Activin–Nodal signaling. Remove either and the cells begin to exit pluripotency within days. The defined E8 formulation that underpins most modern feeder-free culture contains only eight components — DMEM/F12, L-ascorbic acid 2-phosphate, selenium, transferrin, sodium bicarbonate, insulin, FGF2 and TGF-β1 — and that minimal list is itself the evidence: those are the inputs the state actually requires (Chen et al., Nat Methods 2011).

Principle 2 — Dissociation into single cells triggers programmed death

Human pluripotent stem cells survive as colonies. Break a colony into single cells and the majority die — not from mechanical injury, but from a Rho/ROCK-dependent apoptotic program triggered by loss of cell–cell contact. Inhibiting ROCK with Y-27632 suppresses this: in the original report, cloning efficiency of dissociated human ES cells rose from approximately 1% to approximately 27% (Watanabe et al., Nat Biotechnol 2007).

Principle 3 — The culture selects, and it selects against you

Any cell that divides faster or survives passaging better will take over the dish. In pluripotent cultures this produces a characteristic set of recurrent changes. The International Stem Cell Initiative analyzed 125 human ES lines and 11 iPS lines from 38 laboratories and found a progressive tendency to acquire changes on prolonged culture, most commonly affecting chromosomes 1, 12, 17 and 20 — with a minimal amplicon at 20q11.21 containing ID1, BCL2L1 and HM13 appearing in more than 20% of lines (Amps et al., Nat Biotechnol 2011).

Principle 4 — Differentiation is a timed sequence of instructions, not a recipe

Directed differentiation recapitulates development in compressed time. The identity you get is determined by which pathways you block or activate and in what order. The canonical example is dual SMAD inhibition: blocking BMP signaling and TGF-β/Activin–Nodal signaling simultaneously converted more than 80% of human pluripotent cells to neural fate under adherent conditions (Chambers et al., Nat Biotechnol 2009). The same TGF-β pathway that maintains pluripotency in phase 2 must be inhibited to leave it in phase 3.

Principle 5 — Line identity and passage number are experimental variables

Differences between donor individuals, genetic stability and ordinary experimental variability all contribute to iPSC model variation, affecting differentiation potency, cellular heterogeneity, morphology, and transcript and protein abundance (Volpato & Webber, Dis Model Mech 2020). A phenotype observed in one patient line versus one unrelated control line is, statistically, an n of 1 in each arm — the comparison is confounded by the entire genetic background.

How iPSC culture works, step by step

This is how iPSC culture works end to end. Steps 1 and 2 apply only if you are deriving your own lines; most projects start at step 3 with an established line, and some skip to step 8 entirely.

1. Choose a somatic source and a reprogramming method. Dermal fibroblasts, peripheral blood mononuclear cells and urine-derived epithelial cells are the common starting materials; blood and urine are less invasive to collect, fibroblasts are the most forgiving to reprogram. The four-factor set that first reprogrammed adult human fibroblasts — OCT3/4, SOX2, KLF4 and c-MYC — remains the reference point (Takahashi et al., Cell 2007). The delivery method matters more than the source, because it determines whether the resulting line carries a permanent genomic footprint.
   The scienceIntegrating vectors (retro-/lentiviral) insert the reprogramming cassette into the genome, risking insertional mutagenesis and residual transgene expression. Non-integrating approaches avoid this: adenoviral vectors are episomal and are diluted out as cells divide, and Sendai virus and episomal plasmid methods work on the same principle. BioHippo stocks a polycistronic human KLF4-OCT4-SOX2-c-MYC (KOSM) adenovirus in which all four factors are joined by 2A peptides and an IRES for expression from a single promoter, plus a single-factor human OCT4 adenovirus for factor-substitution experiments, within the wider Adenoviral Vectors range. Sendai and episomal reprogramming kits are not in the catalog.
2. Wait for colonies, then pick clonally. Reprogramming is inefficient and asynchronous. Over roughly three to five weeks, a small fraction of transduced cells acquire the compact, tightly packed, high nucleus-to-cytoplasm morphology of a pluripotent colony. Pick individual colonies mechanically and expand them separately — each becomes a clonal line with its own properties.
   The sciencePartially reprogrammed colonies look similar early on but have not activated the endogenous pluripotency network; they fail on passaging or drift back toward the somatic state. The distinguishing test is endogenous OCT4/NANOG expression once the delivered factors have been diluted out — which is why the characterization in step 5 is not optional. Reporter systems that read pluripotency-factor activity, such as an OCT4 reporter lentivirus or a SOX2-OCT4 reporter lentivirus from the Lentiviral Vectors range, let you monitor this live rather than by endpoint staining.
3. Establish feeder-free maintenance: matrix, medium, daily feed. Modern hPSC culture is feeder-free — cells grow on a defined extracellular matrix coating (Matrigel, recombinant vitronectin or laminin-521) in a defined medium, replacing the mouse embryonic fibroblast feeder layers of early protocols. Feed daily. Passage before colonies touch.
   The scienceFeeders supplied both matrix and secreted factors; defined media replace the factors, and the coating replaces the matrix. Vitronectin and laminin-521 are chemically defined and lot-consistent; Matrigel is a tumor-derived extract with substantial lot-to-lot variation, which is a real and underrated source of experimental noise. BioHippo does not stock matrix coatings or hPSC maintenance media — source these from a dedicated stem cell reagent supplier.
4. Passage on a fixed schedule, and choose your dissociation method deliberately. Enzyme-free EDTA passaging lifts cells as small clumps and preserves cell–cell contact, which suits routine maintenance. Enzymatic single-cell dissociation is required for accurate counting, clonal work, sorting and most plate-based assay seeding — and requires ROCK inhibition.
   The scienceClump passaging keeps survival high without a ROCK inhibitor because the apoptotic trigger in Principle 2 is loss of contact, not enzymatic exposure per se. Single-cell passaging gives you a countable, uniform suspension at the cost of that protection — so Y-27632 is added at seeding and withdrawn at the first feed. Do not mix the two methods within an experiment: they produce different attachment kinetics, different colony geometry and different differentiation baselines. EDTA solutions and dissociation enzymes are not in the BioHippo catalog.
5. Verify pluripotency by marker panel, not by morphology. Morphology is a fast daily screen and a poor certificate. Confirm the pluripotent state with a marker panel spanning transcription factors and surface antigens, and confirm the absence of lineage markers.
   The scienceThe core network is OCT4 (POU5F1), SOX2 and NANOG, supported by LIN28A; the standard surface antigens are SSEA-4 and TRA-1-60/TRA-1-81. Nuclear transcription factors are read by immunocytochemistry or intracellular flow; surface antigens by live flow cytometry, which is the more quantitative measure of what fraction of the population is actually pluripotent. Antibodies in catalog: anti-POU5F1/OCT3/OCT4, anti-SOX2, anti-NANOG, anti-LIN28A and anti-SSEA-4 (clone MC813-70). TRA-1-60 and TRA-1-81 antibodies are not currently stocked — you will need a second supplier to complete the surface panel.
6. Monitor genomic stability and sterility on a schedule you set in advance. Karyotype or CNV-array at bank creation, at a defined interval during expansion, and before any experiment that will generate publishable data. Test for mycoplasma monthly. Confirm donor identity by STR profiling.
   The scienceG-banding detects whole-chromosome and large structural changes but will miss the sub-microscopic 20q11.21 gain that is the single most common culture adaptation (Principle 3). SNP array, digital PCR or targeted qPCR for the amplicon covers that blind spot. These are the checks that catch a problem before it invalidates six months of work, and they are also the checks most often skipped. Karyotyping, CNV analysis, STR profiling and mycoplasma detection are not offered in the BioHippo catalog — use a cytogenetics core or a commercial cell line authentication service.
7. Differentiate: exit pluripotency, specify lineage, then mature. Directed differentiation runs in three conceptual stages — exit from pluripotency into a germ layer, patterning to a regional identity, and maturation to a functional cell. Each stage has its own factor set and its own duration, and skipping the maturation stage is why so many iPSC-derived cells behave like fetal rather than adult cells.
   The scienceFor neural lineages, dual SMAD inhibition (BMP block plus TGF-β/Activin block) drives efficient neuroectoderm conversion, after which positional identity is set by Wnt, retinoic acid and SHH gradients and maturation is promoted by Notch inhibition and neurotrophic support. Relevant tool compounds and proteins are linked in Principle 4; the broader Biochemicals and Proteins & Peptides collections hold the wider small-molecule and recombinant factor range. Expect to spend real time optimizing a published protocol for your line — differentiation efficiency is one of the properties that varies most between lines.
8. Characterize the product, not just the process. A differentiation protocol that “worked” is one where you have measured identity, purity and function. Marker expression alone establishes identity; it does not establish that the cells do anything.
   The scienceIdentity by immunocytochemistry or flow for lineage markers, purity as the percentage of the population expressing them, and function by an assay appropriate to the cell type — electrophysiology or calcium imaging for neurons, contractility for cardiomyocytes, secretion for endocrine cells. Viability and metabolic health across the protocol can be tracked with plate-based assays such as Cell Counting Kit-8 or CellQuanti-Blue, which is useful for catching the toxicity of a small molecule in a long protocol before it costs you the whole differentiation.

When to use iPSC culture — and when a simpler model is better

iPSC culture is the correct answer to a narrow class of questions and an expensive wrong answer to a wide class of others. The honest framing is that it buys you two things nothing else provides: human cells of a type you cannot obtain, and a defined, editable genetic background. If your question does not require at least one of those, a primary cell or an immortalized line will almost certainly serve you better.

Use iPSC culture when

• The cell type is inaccessible in humans. Cortical neurons, dopaminergic neurons, motor neurons, cardiomyocytes and hepatocytes cannot be biopsied at scale from living donors. This is the strongest case for iPSC culture and the reason the field exists.
• You need the patient’s genome in the dish. Modeling a mutation in its native genetic background — including polygenic and sporadic disease, where the causal variant is unknown — requires cells derived from that individual.
• You need isogenic comparison. Editing a variant into or out of a single line gives a control that differs at one locus, which is unobtainable with primary tissue from different donors.
• You need a renewable, homogeneous supply for screening. Primary human neurons cannot be expanded; iPSCs can be banked and differentiated repeatedly from the same starting material.
• Species differences invalidate the animal model. Drug metabolism, cardiac ion channel composition and many immune receptors differ enough between rodent and human to make human cells the only defensible system.

Do not use iPSC culture when

• An immortalized line answers the question. Pathway mechanism, protein interaction, reporter assays and most target-validation work do not need pluripotent-derived cells. A validated line from the cell lines range is cheaper, faster and more reproducible.
• Primary cells of that type are commercially available. Endothelial cells, keratinocytes, hepatocytes, PBMCs and many others can be bought as primary human material — with a real donor history and no reprogramming artefacts.
• You need adult-mature phenotypes. iPSC-derived cells are transcriptionally and functionally closer to fetal than adult cells in most lineages. Aging-dependent phenotypes are a known weakness of the system, not a solved problem.
• You need it in under three months. Deriving, characterizing, banking and differentiating a line is a multi-month commitment before the first experiment. If the timeline is short, buy differentiated cells.
• No one in the group has run pluripotent culture before. This is not a technique that transfers from standard cell culture by reading a protocol. Budget for training or a collaboration.

The third option: buy the differentiated cells

The choice is not binary between “culture iPSCs” and “abandon the model.” If you need iPSC-derived cells but not the pluripotent culture itself, you can skip phases 1–3 entirely and start from cryopreserved, characterized, differentiated cells. This removes the two hardest parts of the workflow — maintaining pluripotency and optimizing differentiation — and costs you the ability to make new cell types or edit the genome.

BioHippo’s iPSC-Derived Cells collection covers this route, weighted toward neurological models: cortical neurons, dopaminergic neurons, motor neurons, sensory neurons, astrocytes, neural stem cells and skeletal muscle myoblasts.

The disease-model subset is where this route is most useful, because several mutant lines ship with matched isogenic controls — directly addressing the confound described in Principle 5. TDP-43 motor neuron sets are available as heterozygous, homozygous and isogenic-control trios for M337V HET, M337V HOM and M337V ISO, with comparable sets for Q331K and N352S. SOD1 A4V lines are available as motor neurons, neural stem cells and skeletal myoblasts, and Parkinson’s models include patient-derived sporadic astrocytes and CHCHD2 R145Q dopaminergic neurons.

Decision matrix: matching the route to your research question

Read across from the situation that matches yours. “Full iPSC culture” means running phases 2 and 3 in-house; “buy differentiated” means starting from cryopreserved iPSC-derived cells.

Cardiomyocytes and hepatocytes are named here as the correct route for those questions, but are not currently in the BioHippo iPSC-derived range, which is weighted toward neural and skeletal muscle lineages.

iPSC culture applications across research areas

The iPSC culture applications below are the ones where the technique’s specific advantages — human origin, patient genotype, renewability, editability — do work that no other system can do.

Neurological disease modeling

This is the field’s center of gravity, and for a structural reason: the human brain is the tissue least accessible for biopsy and the one where rodent models diverge most. ALS, Parkinson’s disease, Alzheimer’s disease and frontotemporal dementia are all modeled by differentiating patient-derived lines to the vulnerable cell type and looking for cell-autonomous phenotypes — protein aggregation, mitochondrial dysfunction, altered excitability, selective death. The Neuroscience Cells range and the iPSC-Derived Cells collection both sit in this space.

Isogenic variant-effect studies

Introducing or correcting a single variant in one genetic background converts a correlational observation into a causal test. This is the application where iPSC culture is not merely convenient but methodologically necessary — no primary-cell comparison can control the genome to the same degree.

Drug screening and cardiac/hepatic safety pharmacology

Human iPSC-derived cardiomyocytes and hepatocytes are now used in preclinical safety assessment because species differences in ion channel composition and drug metabolism make rodent data a poor predictor of human risk. Renewability is what makes screening feasible: the same banked line supplies every plate across a campaign.

Developmental biology and organogenesis

Directed differentiation is applied developmental biology run forward in a dish. Because you control the signaling environment at each step, iPSC culture allows patterning questions — how a Wnt or retinoic acid gradient sets positional identity — to be asked in human cells at a resolution not available in a human embryo.

Cell therapy and regenerative medicine research

The long-horizon application: generating transplantable cells from a patient’s own or an HLA-matched genome. Clinically-oriented work imposes requirements the research workflow does not — xeno-free and GMP-grade reagents, rigorous genomic characterization, and documented absence of residual undifferentiated cells. Nothing in this guide, and nothing in the BioHippo catalog, is intended or qualified for clinical use.

Reprogramming and pluripotency mechanism

The reprogramming process itself remains an active research area — factor stoichiometry, chromatin barriers, the role of c-MYC, non-integrating delivery. Work here uses reprogramming vectors as experimental variables rather than as means to an end, which is where the Adenoviral Vectors and Lentiviral Vectors ranges apply.

iPSC culture failure modes and what they are telling you

Most iPSC culture problems present as one of a small number of visual or functional signatures. Read the signature, not the symptom.

Reagents for iPSC culture referenced in this guide

Every item below is live in the BioHippo catalog and is linked at the step where it is used. This is not a complete iPSC culture reagent list — see the scope notes for what is not stocked.

Verify current specifications, pack size and intended-use statements on each product page before purchase. All items are For Research Use Only. Need something not listed here? Request a quote and we will tell you plainly whether we can source it.

iPSC culture FAQ

How many passages can I keep an iPSC line before it is no longer trustworthy?

There is no universal number, which is itself the important answer. Genomic change accumulates progressively rather than at a threshold, and the rate varies by line and by culture conditions. The practical approach is to bank a large stock at low passage, define a working passage window for a given project, verify genomic integrity at the start and end of that window, and return to the bank rather than continuing to passage indefinitely. If growth or differentiation behavior changes noticeably, treat that as a signal to test rather than as good luck.

Do I always need a ROCK inhibitor when passaging iPSCs?

No — only when you generate a single-cell suspension. Enzyme-free clump passaging preserves the cell–cell contacts whose loss triggers the apoptotic program, so it does not require one. You do need it for enzymatic single-cell passaging, for thawing, after sorting, and for clonal isolation. Add it at seeding and remove it at the first medium change; chronic exposure is not a neutral condition and can affect cytoskeletal and differentiation behavior.

Is a normal karyotype sufficient evidence of genomic integrity?

It is necessary but not sufficient. G-banding resolves whole-chromosome and large structural changes, but the most common culture adaptation in human pluripotent cells is a sub-microscopic gain at 20q11.21 that karyotyping will not see. ISSCR standards recommend testing specifically for recurrent CNVs not detectable by karyotype, and for TP53 mutations, when a line shows altered growth or differentiation properties. In practice that means pairing karyotype with an array, digital PCR or targeted qPCR approach for the known hotspots.

Why do my iPSC-derived neurons look fetal rather than adult?

Because they largely are. Directed differentiation recapitulates development, and development does not include aging. Most published protocols produce cells that transcriptionally and functionally resemble fetal or early postnatal stages. Extended maturation, co-culture with astrocytes, and three-dimensional or organoid formats improve this, and progeria-factor or stress-based aging approaches are active research areas — but this remains a genuine limitation of the system rather than a protocol error on your part. If your phenotype of interest depends on cellular age, plan for it explicitly at the design stage.

Should I reprogram my own lines or obtain them from a repository?

Obtain them, unless the genotype you need does not exist. Derivation is a multi-month project that ends with a line you still have to characterize fully, and repositories such as WiCell, EBiSC, Coriell/NIGMS and RIKEN BRC hold thousands of characterized lines including many disease genotypes with consent documentation already in place. Reprogram when the patient genotype is unavailable, when consent for a specific cohort is part of the study design, or when reprogramming itself is the research question.

What is the minimum iPSC characterization I should report in a methods section?

Line identity and source (including repository ID where applicable), passage number at the time of experiment, reprogramming method, culture medium and matrix with supplier and lot where feasible, passaging method, most recent karyotype or CNV result with its date, mycoplasma status, and the pluripotency marker panel used with the fraction of cells positive. For differentiated cells, add the differentiation protocol with its source citation, the differentiation batch, and the identity and purity readouts. This is not excessive: it is the information another lab needs to interpret whether their result agrees with yours.

Can I compare one patient line against one control line?

You can run it, but you should not draw a disease conclusion from it. With one line per arm, the comparison is confounded by the entire genetic background of two individuals, and donor background is known to affect differentiation potency, morphology and expression independently of any disease variant. The two accepted mitigations are isogenic controls — a genome-edited line differing only at the variant — and multiple independent donor lines per arm. Isogenic sets are the stronger design where the variant is known and editable; multiple donors are the practical route for sporadic or polygenic disease.

Non-integrating or integrating reprogramming — does it matter for a research-only project?

Usually yes, even outside a clinical context. Integrating vectors leave a permanent genomic insertion that can disrupt a locus and can be partially reactivated, meaning residual transgene expression may confound differentiation phenotypes. Non-integrating methods — adenoviral, Sendai, episomal, mRNA — avoid this. Integrating vectors remain useful when you want stable, controllable factor expression as a deliberate experimental variable, for example in studies of reprogramming kinetics or factor stoichiometry.

Working out which route fits your model?

Whether you need iPSC-derived cells ready to plate, isogenic disease sets for a variant comparison, or the reprogramming vectors, tool compounds and QC antibodies to run the culture yourself — tell us the cell type and the endpoint, and we will map the catalog against it and be explicit about what we do not stock. Talk to a specialist or browse the iPSC-Derived Cells collection.

References

1. Chen G, Gulbranson DR, Hou Z, et al. Chemically defined conditions for human iPSC derivation and culture. Nat Methods. 2011;8(5):424–429. PMID 21478862 · doi:10.1038/nmeth.1593
2. Watanabe K, Ueno M, Kamiya D, et al. A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat Biotechnol. 2007;25(6):681–686. PMID 17529971 · doi:10.1038/nbt1310
3. International Stem Cell Initiative; Amps K, Andrews PW, Anyfantis G, et al. Screening ethnically diverse human embryonic stem cells identifies a chromosome 20 minimal amplicon conferring growth advantage. Nat Biotechnol. 2011;29(12):1132–1144. PMID 22119741 · doi:10.1038/nbt.2051
4. International Society for Stem Cell Research. Standards for Human Stem Cell Use in Research (2023; Dec 2023 update) — Section 3: Genomic Characterization. isscr.org/basic-research-standards/genomic-characterization
5. Chambers SM, Fasano CA, Papapetrou EP, Tomishima M, Sadelain M, Studer L. Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling. Nat Biotechnol. 2009;27(3):275–280. PMID 19252484 · doi:10.1038/nbt.1529
6. Volpato V, Webber C. Addressing variability in iPSC-derived models of human disease: guidelines to promote reproducibility. Dis Model Mech. 2020;13(1):dmm042317. PMID 31953356 · doi:10.1242/dmm.042317
7. Takahashi K, Tanabe K, Ohnuki M, et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell. 2007;131(5):861–872. PMID 18035408 · doi:10.1016/j.cell.2007.11.019

This guide is general methodological orientation, not a validated protocol for any specific cell line or application. iPSC handling is line-dependent; confirm all conditions against the line’s own documentation and your own endpoint before adopting them. BioHippo does not currently stock hPSC maintenance media, matrix coatings, dissociation reagents, cryopreservation media, mycoplasma detection kits, or karyotyping and STR authentication services; these are named in the text where required so they can be sourced elsewhere. All products referenced are For Research Use Only (RUO) and are not intended for diagnostic or therapeutic use — verify current status and intended-use statements on the product page before purchase.