Fig 2: Hypoxia increases PYCR1-dependent proline synthesis and export from glutamine(A) Intracellular proline abundance in SUM159PT cells is increased with hypoxia (n = 8, presented as mean ± SD). ∗, p < 0.05(B) Diagram of proline synthesis from glucose and glutamine. αKG, α-ketoglutarate; ALDH18A1, aldehyde dehydrogenase 18 family member A1 (pyrroline-5-carboxylate synthase [P5CS]); PYCR1, pyrroline-5-carboxylate reductase 1; PYCR2, pyrroline-5-carboxylate reductase 2.(C) Protein expression of proline synthetic enzymes does not change with decreasing oxygen tension in SUM159PT cells.(D) Extracellular proline abundance is significantly increased in 0.3% O2 in SUM15PT cells (n = 4, presented as mean ± SD). ∗, p < 0.05(E) The ratio of [U-13C]proline to [U-13C]glutamate from [U-13C]glutamine tracing is significantly increased in 1% O2 (n = 4, presented as mean ± SD). ∗∗, p < 0.01(F) Extracellular proline concentration in SUM159PT cells transfected with siPYCR1 is significantly decreased in hypoxia (0.3%) compared with siNT. siPYCR2 does not significantly alter the proline concentration (n = 3, presented as mean ± SD). ∗∗∗∗, p < 0.0001(G) The contribution of glutamine to the total proline pool (%) is significantly reduced with siPYCR1 and not with siPYCR2 (n = 3, presented as mean ± SD). ∗∗∗∗, p < 0.0001(H) In hypoxia (0.3% O2), abundance of [U-13C]proline from [U-13C]glutamine is increased in SUM159PT PYCR1+/+ cells. This increase is not seen in PYCR1−/− cells (n = 3, presented as mean ± SD).(I) Total extracellular proline is increased in hypoxia (0.3%) in PYCR1+/+ cells. This increase is not seen in PYCR1−/− cells (n = 3, presented as mean ± SD). ∗∗, p < 0.01. αKG, α-ketoglutarate; ALDH18A1, aldehyde dehydrogenase 18 family member A1; GSA, glutamate semialdehyde; P5C, pyrroline 5-carboxylate; PYCR1, P5C reductase 1; PYCR2, P5C reductase 2.

Fig 3: PYCR1 activity supports normal TCA cycle activity and proliferation(A) The NAD/NADH ratio is significantly reduced in PYCR1−/− cells (n = 7, presented as mean ± SD). ∗∗, p < 0.01(B) Total extracellular lactate (ion count/protein) is increased in PYCR1−/− cells compared with PYCR1+/+. In hypoxia (0.3%) there is no significant change (n = 4, presented as mean ± SD). ∗, p < 0.05(C) Diagram showing glutamate entry into the TCA cycle, highlighting NAD-requiring steps.(D) Percentage incorporation of [U-13C]glutamine into M + 5 glutamate is unchanged between conditions (n = 4, presented as mean ± SD).(E) Percentage incorporation of [U-13C]glutamine into M + 4 succinate is decreased in the PYCR1−/− cells in hypoxia (n = 4, presented as mean ± SD). ∗∗∗, p < 0.001(F) Percentage incorporation of [U-13C]glutamine into M + 4 malate is reduced in both oxygen tensions in the PYCR1−/− cells (n = 4, presented as mean ± SD). ∗∗, p < 0.01; ∗∗∗∗, p < 0.0001(G) Percentage incorporation of [U-13C]glutamine into M + 4 citrate is reduced in both oxygen tensions in the PYCR1−/− cells and is almost undetectable in 0.3% oxygen (n = 4, presented as mean ± SD). ∗∗∗, p < 0.001; ∗∗∗∗, p < 0.0001(H–J) Cell growth with PYCR1 loss. Statistical analysis shown compares PYCR1+/+ and PYCR1−/− cells without proline. Comparison of growth of either cell type with or without proline was not significant in 21% O2 (H), p < 0.05 in 1% O2 (I), and p < 0.01 in 0.3% O2 (J) in DMEM/Nutrient Mixture F-12 Ham. Growth was not altered by the addition of 2 mM exogenous proline in any oxygen tension (n = 3, presented as mean ± SEM).(K) Cell growth of PYCR1−/− cells is maintained in DMEM flux medium without proline and is not rescued with the addition of exogenous proline (n = 3, presented as mean ± SEM). Statistical analysis shown compares PYCR1+/+ and PYCR1−/− cells without proline. ∗, p < 0.05. Comparison of growth of either cell type with or without proline was not significant.(L) Cell growth of PYCR1−/− cells can be rescued with the addition of 2 mM sodium pyruvate (n = 3, presented as mean ± SD).∗, p < 0.05; ∗∗, p < 0.01; ∗∗∗, p < 0.0001

Fig 4: Loss of PYCR1 in 3D culture leads to redox imbalance and excessive oxygen consumption(A) Volume of PYCR1−/− and PYCR1+/+ spheroids grown on agarose-coated plates (n = 3, presented with Tukey whiskers).(B) Representative images of spheroids from (A). Scale bars, 100 μm.(C) Volume of PYCR1−/− and PYCR1+/+ spheroids at 72 h grown in U-bottomed plates (n = 3, presented with Tukey whiskers).(D) Representative images of spheroids from (C). Scale bars, 400 μm.(E) Hematoxylin and eosin (H&E) staining of sectioned spheroids showing that PYCR1−/− spheroids have an altered morphology, with less densely packed cells and more rounded nuclei. Scale bars, 100 μm.(F) The percentage incorporation of [U-13C]glutamine into M + 4 aspartate is lower in PYCR1−/− spheroids, presented as mean ± SD. ∗, p < 0.05(G) The percentage incorporation of [U-13C]glutamine into M + 4 citrate is lower in PYCR1−/− spheroids (n = 4 [× 24 spheroids], presented as mean ± SD). ∗, p < 0.05(H) Glucose consumption is increased in the PYCR1−/− spheroids while lactate export into the medium is increased (n = 3 [× 24 spheroids], presented as mean ± SD). ∗∗, p < 0.01; ∗∗∗∗, p < 0.0001(I) The ratio extracellular lactate to extracellular pyruvate is higher in PYCR1−/− spheroids (n = 4 [× 24 spheroids], presented as mean ± SD). ∗, p < 0.05(J) PYCR1+/+ and PYCR1−/− spheroids stained for carbonic anhydrase IX expression, as a marker of hypoxic cells. More intense and diffuse staining is seen in the PYCR1−/− spheroids. Scale bars, 70 μm.

Fig 5: PYCR1 loss leads to increased tumor hypoxia(A) Representative images showing pimonidazole staining in xenografts.(B) Representative images showing CA9 staining in xenografts.(C) Representative images showing GLUT1 staining in xenografts.(D) Representative images showing CD31 staining in xenografts.(E) Quantification of pimonidazole staining as percentage of positive cells, presented as mean ± SD. ∗∗∗∗, p < 0.0001. Doxycycline (DOX)-induced shPYCR1 xenografts have significantly higher staining positivity, indicating that PYCR1 loss increases hypoxia.(F) Quantification of CA9 staining as percentage of positive cells, presented as mean ± SD. ∗∗∗∗, p < 0.0001. Doxycycline-induced shPYCR1 xenografts have significantly higher staining positivity.(G) Quantification of GLUT1 staining, as percentage of positive cell, presented as mean ± SD. ∗∗∗, p < 0.001. s. Doxycycline-induced shPYCR1 xenografts have significantly higher staining positivity.(H) Quantification of CD31 staining, presented as mean ± SD. ∗, p < 0.05. Doxycycline-induced shPYCR1 xenografts show a small but statistically significant decrease in staining positivity, suggesting that angiogenesis is impaired by PYCR1 loss.All scale bars represent 250 μm. All experiments: n = 5.