Temperature shift is the single most widely used process lever in commercial CHO fed-batch manufacturing for improving monoclonal antibody titer. The strategy is straightforward: grow cells at 37°C to accumulate biomass, then reduce the temperature to 32–33°C to boost cell-specific productivity and extend culture viability during the production phase. This article explains the mechanism, optimal timing and magnitude, quality implications, and how to design a temperature shift study for your CHO process.
Why Temperature Shift Works
Temperature reduction from 37°C to 32–33°C increases mAb titer primarily by arresting CHO cells in the G1/G0 phase of the cell cycle, redirecting cellular resources from proliferation toward protein synthesis and secretion. The mechanism involves several interconnected effects that collectively favour recombinant protein production over cell division.
At 37°C, CHO cells double every 18–24 hours (specific growth rate μ ≈ 0.029–0.039 h−1). At 32°C, μ drops to 0.005–0.015 h−1, effectively halting division while cells remain metabolically active. This G1/G0 arrest is associated with upregulation of cold-inducible RNA-binding proteins (CIRP and RBM3), increased mRNA stability, and reduced protease activity in the culture supernatant.
The net result: cell-specific productivity (qP) typically rises 1.5–3×, culture viability stays above 90% at day 14 instead of dropping to 70–80%, and the extended viable culture duration increases the integral of viable cell density (IVCD). Since final titer ≈ qP × IVCD, the combined improvement drives a 20–40% titer increase in most CHO mAb processes.
- G1/G0 cell cycle arrest — diverts energy from DNA replication to protein synthesis
- Increased mRNA stability — cold-inducible proteins stabilize transcripts encoding the recombinant product
- Reduced apoptosis — lower metabolic stress delays caspase activation and maintains membrane integrity
- Lower protease activity — reduced cell lysis preserves secreted product in the supernatant
- Improved protein folding — slower ER throughput reduces misfolding and aggregation
Biphasic Culture Timeline
A biphasic temperature strategy divides the fed-batch into two distinct phases: a growth phase at 37°C (days 0–4) to maximise biomass, followed by a production phase at 32–33°C (days 4–14) to maximise specific productivity and culture longevity. The transition point is typically triggered when the viable cell density (VCD) reaches 6–10 × 106 cells/mL in mid-to-late exponential phase.
During the growth phase, cells are in mid-exponential growth with a doubling time of 18–24 hours. The CHO growth curve follows the characteristic number-increase (NI) phase during this period. After the temperature shift, cells transition into a prolonged size-increase (SI) and stationary phase where individual cell volume increases but division rate drops sharply.
The production phase at 32–33°C typically accounts for 70–80% of total mAb produced, despite the lower cell density. This is because the 1.5–3× increase in qP more than compensates for the reduced cell number, and the extended culture duration accumulates more product over 10 days of production compared to 5–7 days in a constant 37°C process that crashes earlier.
Choosing the Right Shift Magnitude
The magnitude of the temperature reduction determines the balance between specific productivity gains and biomass losses. Shifts of 4–5°C (to 32–33°C) consistently outperform both mild shifts (to 35°C) and aggressive shifts (below 30°C) across the published literature for most CHO mAb processes.
| Production Temperature | Growth Rate (μ, h−1) | qP Change | Peak VCD Change | Viability Day 14 | Titer Change |
|---|---|---|---|---|---|
| 37°C (control) | 0.029–0.039 | Baseline | Baseline | 70–80% | Baseline |
| 35°C | 0.020–0.030 | +10–30% | −5–10% | 80–85% | +5–15% |
| 33°C | 0.010–0.020 | +50–100% | −15–25% | 85–92% | +20–35% |
| 32°C | 0.005–0.015 | +80–200% | −25–35% | 88–95% | +25–40% |
| 30°C | 0.002–0.008 | +100–250% | −40–55% | 90–96% | +5–20% |
| 28°C | ∼0 | +150–300% | −55–70% | >95% | −10–+5% |
At 35°C, the growth rate reduction is modest and qP improvement is marginal. At 28°C, growth arrest is essentially complete and qP is maximised, but the severe loss of biomass reduces IVCD enough to offset the productivity gain. The sweet spot at 32–33°C balances qP improvement against sufficient residual growth to maintain a productive culture.
Minor temperature differences (even 1–1.5°C) have measurable effects on both titer and product quality. This means bioreactor temperature control must be precise during the production phase — PID loop tuning and jacket temperature control are critical, especially at manufacturing scale where thermal mass creates response lag. Use the Heat Transfer Calculator to evaluate cooling capacity requirements at your target scale.
Optimal Shift Timing and Cell Density Triggers
The temperature shift should be applied in mid-to-late exponential phase, typically on day 3–5 of a 14-day fed-batch when VCD reaches 6–10 × 106 cells/mL. This timing maximises the IVCD by allowing sufficient biomass accumulation before switching to the high-qP production phase.
Two approaches are commonly used to trigger the shift:
- Fixed-day trigger — shift on a pre-determined day (e.g., day 4). Simple and reproducible, suitable when clone growth kinetics are consistent across batches. Used in most commercial GMP processes.
- VCD-based trigger — shift when VCD crosses a threshold (e.g., 8 × 106 cells/mL). Adapts to batch-to-batch variability in inoculum quality and growth rate. More robust in development, but adds a process decision point that requires justification in regulatory filings.
The lactate accumulation profile provides an additional timing signal. Ideally, the temperature shift coincides with or slightly precedes the lactate metabolic shift, where cells transition from lactate production to consumption. This synchronises the growth-to-production transition across metabolic and cell-cycle axes.
Impact on Product Quality Attributes
Temperature shift affects multiple critical quality attributes (CQAs) beyond titer. Understanding these effects is essential for developing a strategy that simultaneously optimises both productivity and product quality, which is a regulatory expectation under ICH Q8 design-space principles.
| Quality Attribute | Direction at 32–33°C | Mechanism | Typical Magnitude |
|---|---|---|---|
| Galactosylation (G1F + G2F) | Increase | Longer Golgi residence time → more complete β4GalT processing | +10–25 percentage points |
| High mannose (Man5) | Decrease | Reduced ER stress and more efficient Golgi processing | −2–5 percentage points |
| Sialylation | Slight increase | Extended Golgi processing and sialyltransferase activity | +2–8 percentage points |
| Aggregation (HMW) | Decrease | Improved folding efficiency, reduced ER stress, lower protease release | −0.5–2% |
| Charge variants (acidic) | Variable | Product-dependent: deamidation may decrease, sialylation increase | Cell-line specific |
| Product-related impurities | Decrease | Less cell lysis → lower HCP, DNA, and protease content in harvest | −20–50% HCP |
The glycosylation control implications are particularly significant for biosimilar development, where the glycan profile must match the innovator product. Temperature is the most powerful single lever for tuning galactosylation: a shift from 37°C to 32°C can increase galactosylation by 10–25 percentage points. Combined with manganese (5–20 μM MnCl2) and galactose (5–20 mM) supplementation in the feed, fine-tuning of the G0F/G1F/G2F distribution is achievable without affecting titer.
The reduction in aggregation at lower temperature is driven by slower ER throughput allowing more complete disulphide bond formation and chaperone-assisted folding. For aggregation-prone molecules, temperature shift is often the first intervention explored before resorting to media or feed modifications.
CHO Troubleshooter
Diagnose low viability, poor titer, or abnormal metabolite profiles in your CHO fed-batch process.
Triphasic and Advanced Strategies
Triphasic temperature strategies apply a second temperature reduction partway through the production phase, typically from 33°C to 31°C around day 7–8 when viability begins to decline. This second shift extends the productive culture period by an additional 2–3 days and can add 10–15% titer over a biphasic approach.
Other advanced approaches include:
- Gradual ramp-down — a continuous temperature decrease of 0.5°C/day from day 3 to day 10. More complex to implement but avoids the metabolic shock of a step change. Reported to reduce lactate spikes after the shift.
- VCD-feedback control — temperature is dynamically adjusted based on real-time VCD from a capacitance or optical biomass sensor. When VCD exceeds a setpoint, temperature is lowered. This creates a self-regulating system that adapts to clone-to-clone variability.
- Combined temperature + pH shift — shifting pH from 7.0 to 6.8–6.9 simultaneously with the temperature reduction enhances the productivity effect. The combined parameter shift can increase titer by 30–50% over constant conditions.
Heat Transfer Calculator
Evaluate jacket cooling capacity and temperature control dynamics for your bioreactor scale and temperature profile.
Worked Example: Designing a Temperature Shift Study
A systematic temperature shift study for a new CHO mAb clone requires testing at least 3 production temperatures at 2–3 shift timings in a factorial or partial-factorial design. Below is a practical worked example for a typical screening campaign.
Worked Example — Biphasic temperature shift DOE
Objective: Determine the optimal production-phase temperature and shift timing for a CHO-K1 GS clone producing an IgG1 mAb in a 14-day fed-batch at 5 L scale.
Factors:
- Production temperature: 31°C, 32.5°C, 34°C (3 levels)
- Shift day: day 3, day 5, day 7 (3 levels)
- Control: constant 37°C (1 run)
Design: Full factorial (3 × 3) + 1 control = 10 conditions, duplicated = 20 bioreactor runs.
Responses measured:
- Final titer (g/L) at day 14
- Peak VCD (× 106 cells/mL)
- IVCD (× 106 cell·days/mL) = Σ(VCDi × Δti)
- qP (pg/cell/day) = Δtiter / IVCD
- Viability at day 14 (%)
- Glycan profile (G0F%, G1F%, G2F%, Man5%)
Calculation — IVCD for optimal condition (32.5°C shift on day 4):
Growth phase (day 0–4): average VCD = 5.0 × 106/mL × 4 days = 20 × 106 cell·days/mL
Production phase (day 4–14): average VCD = 8.5 × 106/mL × 10 days = 85 × 106 cell·days/mL
Total IVCD = 20 + 85 = 105 × 106 cell·days/mL
With qP = 35 pg/cell/day at 32.5°C:
Estimated titer = qP × IVCD = 35 × 10−12 g/cell/day × 105 × 106 cells·days/mL
= 35 × 105 × 10−6 g/mL = 3.675 × 10−3 g/mL = 3.7 g/L
Compare with constant 37°C control: qP ≈ 18 pg/cell/day, IVCD ≈ 120 × 106 cell·days/mL (higher peak VCD but shorter viable period), estimated titer = 18 × 120 × 10−6 = 2.16 × 10−3 g/mL = 2.2 g/L. The 32.5°C shift yields +68% titer in this example.
Use a DOE approach to efficiently screen the temperature × timing space. A definitive screening design (DSD) with 3 centre points requires only 9 runs to identify the main effects and two-factor interactions, reducing the bioreactor-time investment compared to a full factorial.
Perfusion Calculator
Model continuous perfusion processes as an alternative to fed-batch — compare steady-state productivity and media consumption.
Frequently Asked Questions
What temperature should I shift to in CHO fed-batch culture?
Most commercial CHO fed-batch processes shift from 37°C to 32–33°C. A shift to 32°C typically increases final titer by 20–40% compared to a constant 37°C control. Shifts below 30°C arrest growth too aggressively and can reduce overall IVCD, while shifts to only 35°C often have minimal impact on specific productivity. The optimal temperature is cell-line and product specific and should be determined experimentally.
When should I apply the temperature shift in a CHO fed-batch process?
Apply the temperature shift on day 3–5 of a 14-day fed-batch, once cells reach 6–10 × 106 cells/mL in mid-to-late exponential phase. Shifting too early limits biomass accumulation and reduces IVCD. Shifting too late misses the window to extend viability and boost specific productivity during the production phase.
How does temperature shift affect mAb glycosylation?
Lowering culture temperature from 37°C to 31–33°C increases galactosylation by 10–25 percentage points, shifting the glycan profile from G0F-dominant toward higher G1F and G2F. This occurs because reduced growth rate extends protein residence time in the Golgi, allowing more complete processing by galactosyltransferases. The effect on high mannose and sialylation varies by cell line.
Can I use a triphasic temperature strategy instead of biphasic?
Yes. Triphasic strategies (e.g., 37°C → 33°C on day 4 → 31°C on day 8) can add 10–15% titer over biphasic approaches by further extending culture longevity. They add process complexity and require tighter control, so most commercial processes use biphasic unless the quality profile specifically benefits from the second shift.
Does temperature shift affect cell-specific productivity (qP)?
Yes. Temperature reduction from 37°C to 32–33°C increases qP by 1.5–3×. The increase is associated with G1/G0 cell cycle arrest, increased mRNA stability, reduced metabolic burden from growth, and upregulation of ER/Golgi processing. The net effect on titer depends on the balance between qP increase and VCD reduction.
Related Tools
- CHO Troubleshooter — diagnose low viability, abnormal metabolite profiles, and poor titer in CHO fed-batch culture
- Heat Transfer Calculator — evaluate jacket cooling capacity and temperature control dynamics at manufacturing scale
- Perfusion Calculator — model continuous perfusion as an alternative strategy to fed-batch for CHO mAb production
References
- Yoon SK, Song JY, Lee GM. Effect of low culture temperature on specific productivity, transcription level, and heterogeneity of erythropoietin in Chinese hamster ovary cells. Biotechnology and Bioengineering. 2003;82(3):289-298. doi:10.1002/bit.10566
- Kaufmann H, Mazur X, Fussenegger M, Bailey JE. Influence of low temperature on productivity, proteome and protein phosphorylation of CHO cells. Biotechnology and Bioengineering. 1999;63(5):573-582. doi:10.1002/(SICI)1097-0290(19990605)63:5<573::AID-BIT7>3.0.CO;2-Y
- Mason M, Sweeney B, Cain K, Stephens P, Sharfstein ST. Identifying bottlenecks in transient and stable production of recombinant monoclonal-antibody sequence variants in Chinese hamster ovary cells. Biotechnology Progress. 2012;28(3):846-855. doi:10.1002/btpr.1542
- McHugh KP, Xu J, Aron KL, Borys MC, Li ZJ. Effective temperature shift strategy development and scale confirmation for simultaneous optimization of protein productivity and quality in Chinese hamster ovary cells. Biotechnology Progress. 2020;36(3):e2959. doi:10.1002/btpr.2959
- Xu WJ, Lin Y, Mi CL, Pang JY, Wang TY. Progress in fed-batch culture for recombinant protein production in CHO cells. Applied Microbiology and Biotechnology. 2023;107(4):1063-1075. doi:10.1007/s00253-022-12342-x