*2.1. Population Characteristics*

Table 1 summarizes the maternal characteristics and pregnancy outcomes for the normotensive control and early-onset PE cohorts. Women with PE were hypertensive (162.3 ± 11.8/102.9 ± 5.9 mmHg) with confirmed proteinuria. These women were younger than controls (*p* < 0.01) and more likely to be nulliparous (*p* < 0.01). There was no significant di fference in gestational age between control and PE

cohorts. However, infants born from pregnancies complicated by PE had lower birth (*p* < 0.05) and placenta (*p* < 0.05) weights compared to controls.


**Table 1.** Population Characteristics.

1 Mean (standard deviation); 2 Number (%) total number; 3 Median (interquartile range); NS: not significant.

### *2.2. Placental Creatine and GAA Content*

Total creatine content of PE placentae was significantly higher (+38%) than controls (Figure 1A; *p* < 0.01). Placental content of GAA (creatine precursor) was not significantly different between PE and control cohorts (Figure 1B). Baby sex had no effect on placental creatine or GAA content in either control or PE pregnancies.

**Figure 1.** Placental Creatine and Creatine Precursor Guanidinoacetate (GAA) Content. Total creatine (**A**) and GAA (**B**) content of *n* = 19 healthy control (closed circle) and *n* = 19 PE (open circle) placentae. Two sample t-tests were used for statistical comparison. Data are present means ± SD. \*\* *p* < 0.01.

#### *2.3. Creatine Synthesis and Transport Genes and Proteins*

*GATM* and *GAMT* mRNA expression were up-regulated 2-fold in PE placentae compared to controls (Figure 2A *p* < 0.001 and Figure 2B *p* < 0.001, respectively). We also observed a two-fold increase in mRNA expression for the creatine transporter (*SLC6A8*) and a 4-fold increase in cytosolic *BBCK* mRNA expression in PE placentae compared to controls (Figure 2C *p* = 0.021 and Figure 2D *p* < 0.001, respectively). Conversely, mitochondrial creatine kinase (*CKMT1A*) mRNA expression was significantly decreased in PE placentae (Figure 2E *p* < 0.01). Despite changes in mRNA expression of the creatine synthesizing enzymes and creatine kinases, there were no significant differences in protein abundance of AGAT, GAMT, BBCK or CKMT1A between the PE and control placentae (Figure 3, blots available in Supplementary Information as Figures S1–S4). Baby sex had no impact on placental gene and protein expression in either control or PE pregnancies.

**Figure 2.** Placental Gene Expression of the Creatine Synthesizing Enzymes, the Creatine Transporter and Kinases. Creatine synthesizing enzymes *GATM* gene that codes for AGAT (**A**) and *GAMT* (**B**), the creatine transporter *SLC6A8* (**C**), cytosolic creatine kinase *BBCK* (**D**) and mitochondrial creatine kinase *CKMT1A* (**E**). *n* = 20 normotensive control (closed circle) and *n* = 20 PE (open circle). Data are presented relative to the control cohort. Values are mean ± SD. Wilcoxon Rank Sum was used for statistical comparison. Significance was set at *p* ≤ 0.025, following a Benjamini-Hochberg adjustment for false-positives.

**Figure 3.** Placental Protein Expression of the Creatine Synthesizing Enzymes and Kinases. Creatine synthesis enzymes AGAT (**A**) and GAMT (**B**), cytosolic creatine kinase BBCK (**C**) and mitochondrial creatine kinase CKMT1A (**D**). *n* = 19 normotensive control (closed circle) and *n* = 19 PE (open circle). Data were normalized to total protein and a control sample run across blots. Data are expressed in arbitrary units (A.U.) relative to the control cohort. Wilcoxon Rank Sum was used for statistical comparison. Significance was set at *p* ≤ 0.025, following a Benjamini-Hochberg adjustment for false-positives. NS: not significant.

#### *2.4. Correlations between Creatine Metabolism and Pregnancy Outcomes*

We explored associations between placental metabolite content, gene and protein expression, and then these laboratory measures with maternal characteristics (age, parity, mode of delivery) and pregnancy outcomes (baby sex, gestational age, birth weight and placental weight). There was a significant decrease in placental *GATM* mRNA expression with advancing gestational age (Figure 4A, *r* = −0.681; *p* < 0.001) and birth weight (Figure 4B, *r* = −0.610; *p* < 0.001) in controls. These gene

associations were absent in the PE cohort (Figure 4C,D). Similar correlations were observed with placental GAA, with tissue content declining with advancing gestational age (Figure 4E, *r* = −0.720; *p* < 0.002) and birth weight (Figure 4F,*r* = −0.687; *p* < 0.003) in normotensive controls. These associations were again absent in the PE cohort (Figure 4G,H). Placental creatine content was not significantly associated with any other maternal characteristics or pregnancy outcomes. Nor were mRNA expression patterns of *GAMT, SLC6A8, BBCK, CKMT1A* or AGAT, GAMT, BBCK or CKMT1A protein abundance.

**Figure 4.** Associations (r) between placental *GATM* mRNA and gestational age (**A**) and birth weight (**B**) in the control cohort; placental *GATM* mRNA and gestational age (**C**) and birth weight (**D**) in the PE cohort; placental GAA content and gestational age (**E**) and birth weight (**F**) in the control cohort and GAA content and gestational age (**G**) and birth weight (**H**) in the PE cohort. *r* values were generated using Spearman's correlation coefficient. Normotensive placenta data are represented by closed circles and PE placentae open circles. Significance was set at *p* ≤ 0.004, following a Benjamini-Hochberg adjustment for false-positives.
