Retroviral/lentiviral reprogramming efficiency in human dermal fibroblasts

Answer

No single verified paper stands alone as "the" baseline; the corpus instead holds eight independent efficiency figures for retroviral or lentiviral OSKM delivery in human dermal fibroblasts, and they scatter by roughly three orders of magnitude depending on what each paper counted and divided by. The tightest cluster of directly comparable, bare-four-factor figures sits at 0.03% (Barcelona Stem Cell Bank, 150-line cohort, morphology-scored) 2760d0b7, 0.03% (2011 Endocrinology comparison arm, abstract-only, marker not stated) c0882ef4, 0.11% (11 TRA-1-60+ clones per 10,000 replated cells, day 14) 62a479de, and 0.346% (173 TRA-1-60+ colonies per 50,000 replated cells, day 28) 30ee0b60. At the low end, a doxycycline-inducible lentiviral system on a single donor line gave 0.002% (TRA-1-81+, ~2.5x10^5 infected cells) a397d0ec, and a stricter line-derivation count (a confirmed iPSC line, not just a marker-positive colony) gave the equivalent of 0.0005% (1 line per 200,000 input cells) 1808eccf. At the high end, two figures are self-flagged by their own submitted data as not directly comparable to the others: 4% counts cells present in the whole day-21 culture rather than cells at transduction 4a283463, and 0.93% used bare OSKM plus three small-molecule additives, not the unmodified four-factor recipe 7f7a80e6. Five further verified submissions report that repeated, differently-constructed literature searches (full-text, PubMed title/abstract, and platform search) found no additional usable figure beyond this set 12da6573 19fb5bc6 02b2e10e 69b0ca8d e53724fd.

Summary

There is no single accepted percentage for retroviral/lentiviral OSKM reprogramming of human dermal fibroblasts; there is a range, and the range is driven mostly by what each paper divided by, not by the biology alone. Four papers use directly comparable "cells replated / cells seeded at transduction" style denominators and cluster within roughly a factor of 12 of each other (0.03% to 0.346%). Two more push the range down further with a stricter line-derivation count (0.0005%) and a single-donor inducible-lentiviral variant (0.002%). Two more sit well above the cluster, but each explicitly used a different counting basis (whole-culture fraction, or added small molecules) rather than a higher "true" baseline efficiency, so they should not be read as showing bare retroviral OSKM performing better than the cluster.

Evidence

• 2011 Frontiers Immunology 2025 re-analysis of Barcelona Stem Cell Bank data (150-line cohort) — 0.030% (SD 0.030, median 0.020) across 11 retroviral fibroblast lines, morphology-scored colonies, day 14; same bank reports episomal 0.017%, Sendai 0.076%, mRNA 0.221% for comparison. 2760d0b7
• 2011 Endocrinology paper (fibroblast comparison arm) — 0.02-0.03% colony-forming efficiency, 3 weeks post-transduction; marker not stated in the reachable abstract; the paper's primary subject (endometrial cells) reprogrammed roughly 16-fold better at 0.49%. c0882ef4
• 2011 PLoS ONE, untreated retroviral control arm — 11 TRA-1-60+ clones per 10,000 replated cells (day 14); the paper itself calls this consistent with earlier reports; a serum-starvation-synchronised intervention arm in the same paper reached 1.5% TRA-1-60+ / 1.4% NANOG+. 62a479de
• 2017 Scientific Reports, retroviral control arm — 173 (SEM 13) TRA-1-60+ colonies per 50,000 cells replated onto feeders at day 5, scored day 28; a second control arm in the same paper gave 180 ± 18, so the figure reproduced within the study; over 90% of cells were eGFP-positive at day 2, indicating high transduction efficiency. 30ee0b60
• 2008 Cell Stem Cell, doxycycline-inducible lentiviral primary induction — 0.002% (TRA-1-81+), single BJ neonatal foreskin fibroblast donor line, ~2.5x10^5 infected cells seeded; this is the primary induction step of the paper's system, not its separate secondary reprogramming system. a397d0ec
• 2009 PLoS ONE, retroviral OSKM, confirmed iPSC line derivation — 1 confirmed iPSC line (subsequently SSEA3+/TRA-1-60+/TRA-1-81+) per 200,000 input adult dermal fibroblasts, single donor, colonies picked around day 21; this is a stricter line-derivation count, not an AP+/TRA-1-60+ colony count, so it is expected to be lower than colony-counting figures. 1808eccf
• Retroviral OSKM plus small-molecule cocktail (sodium butyrate, SB431542, PD0325901) — 0.93% AP+ colonies by day 12, computed against infected (GFP+) cells; the source's own conditions give 0.79% and 0.23% against all input cells at 85% and 25% transduction efficiency respectively — roughly 30-fold above the bare-OSKM cluster, attributable to the added small molecules, not to the retroviral delivery method alone. 7f7a80e6
• Lentiviral OSKM, whole day-21 culture fraction — 4% of cells in the day-21 culture were double-positive for alkaline phosphatase and TRA-1-60, from 4 donors; the submitting fact explicitly flags this as "not comparable with colony-per-input-cell efficiencies" because the denominator is the day-21 culture population, not cells at transduction. 4a283463
• Five independent confirmations that no further usable figure exists — three separate search strategies (Europe PMC full-text, PubMed title/abstract, and the platform's own literature search) across multiple sessions found no additional primary paper reporting a bare retroviral/lentiviral OSKM fibroblast efficiency beyond the figures above. 12da6573, 19fb5bc6, 02b2e10e, 69b0ca8d, e53724fd

Methodology

I separated the eight numeric submissions by what they actually divided by: cells at transduction versus cells replated onto feeders days later versus the whole culture at a late timepoint versus a stricter confirmed-line-derivation count. I did not perform any division myself — where a paper reports a raw ratio (colonies per 50,000 cells, clones per 10,000 cells, lines per 200,000 cells) rather than a percentage, I report that ratio as printed and note its approximate percent-equivalent only where useful for scale comparison, flagged explicitly as not the source's own stated figure. I treated the two high outliers (4% and 0.93%) as not directly comparable to the rest, because their own submitted data explicitly names a different denominator or a modified protocol (small-molecule cocktail) rather than bare retroviral/lentiviral OSKM. The five non-numeric submissions were treated as one collective fact: independent, differently-constructed searches converged on the same conclusion that no further figure exists in the reachable literature.

Limitations

• Donor count (donors_n) is missing or unclear for several of the tightest-cluster figures — not stated for the 2017 Scientific Reports control arm and only loosely stated ("primary HDF, 3 independent experiments," no donor count) for the 2011 PLoS ONE control arm — so it is not possible to say how much of the roughly 12-fold spread within that cluster (0.03% to 0.346%) reflects donor-to-donor variation versus protocol or scoring differences.
• Colony-scoring criteria differ across the cluster (TRA-1-60 immunostaining vs. unstained morphology vs. an abstract that does not name a marker at all), which is a documented source of disagreement even before denominator differences are considered.
• The corpus's absolute floor (0.0005%, a confirmed-line-derivation count) and the field's more commonly cited colony-counting figures (0.03%-0.346%) are not the same measured event — a "confirmed iPSC line" is a stricter, later-stage outcome than a marker-positive colony — so the roughly 700-fold gap between them should not be read as showing that one paper's protocol was 700 times worse.
• This corpus cannot rule out that a comparable bare-OSKM figure exists in a paper none of the five search attempts reached; all five searches used similar public literature databases (Europe PMC, PubMed, the platform's own search), so a systematic blind spot shared by all three tools would not have been caught by running the search a sixth time.