*3.5. PP13 Polymorphism and Preeclampsia*

Polymorphic variants of PP13 have been identified [38,66], and three of them are important indicators of a high risk to develop preeclampsia:

(1) The "truncated" variant is a deletion of thymidine in position 221 of the open reading frame of exon 3 [85,86]. It was discovered among black and colored pregnan<sup>t</sup> women in a Cohort of Cape Town, South African cohort [85]. It is associated with the development of an earlier stop codon coupled to a shorter PP13 variant ("truncated" or "delT221") [38]. The shorter delT221 variant is lacking the entire exon 4 and part of exon 3 [38] (Figure 1). Hence, delT221 is missing 2 of the amino-acids involved in the carbohydrate recognition domain (CRD), and two additional amino acids supporting carbohydrate binding [6,31]. Having this mutation in a heterozygous form is an e ffective predictor of severe early preeclampsia with 89% positive predictive value. Treatment of human leukocytes derived of the maternal decidua with the wild type of recombinant PP13 but not with the truncated PP13 induced apoptosis [31,43]. From this data it was speculated that one role of PP13 in pregnancy is to render the mother immune-tolerant to pregnancy. The immune tolerance is reached by binding of PP13 via the CRD to glycoproteins and glycolipids. Indeed, pregnancies carrying the homozygous DelT221 mutation are rejected by the mother and are not viable [86].

(2) The promoter variant. The -98 (A/C) promoter genotype displays three genotypes: the "A/A" genotype (homozygous to the adenosine nucleotide), the "C/C" genotype (homozygous to cytosine), or the "A/C" genotype (heterozygous form). In a South African as well as a London cohort of pregnan<sup>t</sup> women, the A/A genotype was found to be associated with decreased expression of PP13 compared to the level of PP13 expression with either A/C or C/C genotypes in the -98 position [51,87]. The reduced expression was contributed in part by the impaired ability of the transcription factor TFAP2A to induce PP13 expression with the A/A genotype [51]. Accordingly, carriers of the A/A variant had an adjusted odds ratio of 3.68 to develop preeclampsia, while the C/C or the A/C genotypes rendered protection from developing preeclampsia. Combining the A/A genotype as a risk factor together with black ethnicity, history of previous preeclampsia, obesity (BMI > 37), and being at advanced maternal age provided an adjusted odds ratio of 14.0 and 7.0, respectively, for developing term or all preeclampsia cases [51].

(3) The Dex-2 variant. Recently, we were able to molecularly engineer a third molecular variant of PP13 that was denoted Dex-2. This mutant completely lacks the second exon (Figure 1). This additional natural PP13 variant was initially isolated in Israel while cloning PP13 DNA from a genomic library. The mutant clones were isolated from a placenta obtained after delivery from a woman with preeclampsia combined with FGR [55]. Burger et al. [55] have shown that PP13 which was isolated from a placenta of preeclampsia with FGR was inferior in inducing the liberation of free fatty acids from trophoblast membranes, and in causing the elevated release of prostaglandins. Further analysis of this mutant is warranted.

In summary, there may well be a link between reduced levels of PP13 during the first trimester of human pregnancy and the elevated risk for a subsequent development of preeclampsia. Preeclampsia patients may be a target population to evaluate if nourishing with the wild-type, full length PP13 can be used as a therapeutic tool to fight preeclampsia.

### *3.6. PP13 and Immune Tolerance*

The syncytiotrophoblast secretes/releases PP13 from the first trimester and the protein reaches the decidua either via di ffusion or via the maternal circulation, coinciding with the time of early trophoblast invasion. Kliman et al. [43] have shown the formation of PP13 aggregates closer to areas with increased apoptosis of various maternal immune cells. Killing these cells could enable us to promote extravillous trophoblast invasion of the uterine wall. In this manner, PP13 might serve to establish a decoy inflammatory response, sequestering maternal immune cells away from the site of extravillous trophoblast invading other sites of the uterine wall [31,39,43]. Accordingly, it was proposed that PP13 contributes to the immune tolerance of the mother to the invading trophoblasts. Having low levels of PP13 and/or having a mutated variant may decrease the level of PP13 secretion, thereby contributing to impaired placentation.

### *3.7. PP13 Replenishing Studies in Animals*

The uteroplacental circulation undergoes massive changes during pregnancy, resulting in a vascular system that is directing 20% of the total cardiac output to the uterine vascular bed. This results in more than a ten-fold increase in blood flow over the level present in the non-pregnan<sup>t</sup> state [88]. Since in normal pregnancy there is only a small drop in blood pressure, it is necessary to gain uterine hemodynamic changes by uterine blood vessel expansion and reduced uterine vascular resistance [89]. In pregnancy, extravillous trophoblasts invade all types of luminal structures in the placental bed [90,91]. One of their major targets are spiral arteries and their adjacent stroma. The endoarterial trophoblast subpopulation [92] replaces and reorganizes the vascular smooth muscle and endothelial layers, resulting in the formation of low-resistance vessels that can accommodate a highly increased blood volume flowing towards the placenta [93]. These altered vessels are almost independent of maternal vasoconstriction through a lack of smooth muscle cells [89,94].

Through the invasive processes of the extravillous trophoblast, the vessels towards the intervillous space of the placenta (spiral arteries) and those draining blood back into the maternal system (uteroplacental veins) are connected to the placenta, resulting in a placental blood flow to su fficiently supply the placenta and the growing fetus with nutrients and oxygen [89,91]. This hemochorial type of placentation is present in mammals such as humans, higher order primates, rabbits, guinea pigs, mice, and rats [95–97]. At term there are around 200 spiral arteries opening towards the intervillous space, while the blood flow in the uterine artery is increased in volume with reduced velocity [96]. Impaired trophoblast invasion into spiral arteries results in higher blood flow velocities into the intervillous space of the placenta and thus damage of the fragile villous trees. [98–101].

PP13 appears to have an important role in these hemodynamic changes by facilitating expansion of the uterine vascular system during pregnancy to accommodate the increase of blood flow through the uterus and thus the placenta during pregnancy. The following in vivo data have been obtained using PP13 administration in di fferent animal models:

• Initially, a single PP13 dosage injected intravenously into gravid rats and rabbits resulted in a reversible ~30% reduction in blood pressure [102].

• In a second set of experiments, peristaltic pumps were implanted into gravid rats for a slow release of PP13 for 4 to 7 days from day 15 [102], or from day 8 of pregnancy [103]. rPP13 (compared to saline control) reversibly reduced blood pressure until the pumps released all their content. At delivery, 5 to 7 days after the active release of PP13 was over, treated animals had larger placentas and pups. Both the wild type rPP13 and the truncated variant DelT221 were e ffective in reducing blood pressure, but the truncated variant failed to sustain uterine artery expansion until the time of delivery [103].

• Isolated uterine mesenteric arteries from both mid-pregnant and non-pregnan<sup>t</sup> rats were placed in arteriographs to measure their diameters and pressure in response to drug perfusion [104]. Uterine arteries of both pregnan<sup>t</sup> and non-pregnan<sup>t</sup> rats were dilated in a dose dependent manner with increasing concentrations of PP13. Half-maximal vasodilation of isolated arteries (EC50) was achieved at a concentration of 1pM PP13 (blood level of pregnan<sup>t</sup> women). The e ffect was mediated by the endothelial layer, since stripping the vessels o ff the endothelial layer prohibited blood vessel expansion by PP13. Pharmacological analysis of the signaling pathways revealed that the vasodilation was mediated through signaling of the endothelial nitric oxide synthase (eNOS) and prostaglandin type 2 pathways [104].

• An additional study was performed with non-pregnan<sup>t</sup> rats. Again, surgically implanted pumps released a constant dose of PP13 (rPP13 or His-PP13 variants) or saline over seven days. Some animals were sacrificed immediately after the end of PP13 release (on day 7), while others were sacrificed 6 days later (day 13) to compare the short and long-term impacts of PP13 on vessel growth and size. Both uterine veins and arteries were significantly expanded by rPP13 with a more pronounced effect after 13 days compared to the corresponding vessels after seven days. The long-term e ffect of treatment by rPP13 was more pronounced in the veins compared to the corresponding arteries. His-PP13 also expanded the blood vessels but the e ffect remained similar between 7 and 13 days, most

likely since His-tag PP13 has only a monomeric form. This molecular variant does not turn into the natural configuration of a homo-dimer. It is estimated that to exert the structurally stable vascular expansion that is developed with the non His-tag protein, a molecular variant that forms a homo-dimer is required [105].

In conclusion, PP13 appears to play a key role in the remodeling of uterine arteries and veins during pregnancy, facilitating the adjustment of blood flow to and from the placenta. This way, PP13 adapts the uterus to provide increased but slower blood flow towards the placenta and back into the maternal system, necessary for normal pregnancy. PP13 acts via the NO and prostaglandin signaling pathways to provide oxygen and nutrients to the growing fetus (Figure 3).

**Figure 3.** PP13 priming of maternal blood vessels. The scheme displays a comprehensive model of the PP13 effects on the vascular system of the mother. PP13 is released from the syncytiotrophoblast and enters the maternal blood system where it has different effects. The red arrow shows flow of maternal blood into the placenta via invaded spiral arteries, while the blue arrows indicate flow of maternal blood back from the placenta into the maternal vascular system via invaded uterine veins.

#### *3.8. Modeling the Role of PP13 in Pregnancy*

The PP13 molecule primes blood vessel expansion to adapt the uterine vascular system to supply oxygen and nutrients to the growing fetus. The effect supports the development of larger placentas and pups, as shown in a rat model [103]. The effect involves a chain of reactions, starting from a physiological effect that involves the endothelial layer through the e-NOS and prostaglandin signaling pathways [103], and continuing through structural stabilization of the surrounding components of the connective tissue around the blood vessels. Connective tissue stabilization requires the CRD component of the PP13 molecules that crosslinks between the endothelial layer and the connective tissue (Figure 4). Finally, PP13 acts as a decoy to attract maternal immune cells and thereby enabling the invasion of extravillous trophoblast into blood vessels [43].

Based on all the above we propose a targeted PP13 therapy to fight preeclampsia in patients with impaired PP13 and high risk to develop preeclampsia [106]. These patients could receive PP13 as a nourishing drug to support uterine vessel expansion and stabilization of blood supply during pregnancy. We are conducting preclinical studies and plan to evaluate the potential clinical impact testing animal models of preeclampsia to explore this hypothesis.

Considering the multifaceted nature of preeclampsia [58–60], the development of PP13 as a novel biological therapy to fight preeclampsia is now evaluated in certain animal models to provide a proof of concept. Among animal models, we plan to test (1) the reduced uterine placenta perfusion (RUPP model) in rats [107], (2) the transgenic mouse model of STOX-1 [108], and (3) the Baboon uteroplacental

ischemia model [109], all identified as important model systems to evaluate novel drugs to fight preeclampsia [110,111].

**Figure 4.** The effect of PP13 on the various layers of blood vessels. PP13 acts on the endothelial layer of the blood vessels and causes vasodilation by muscle relaxation through the signaling pathways of eNOS and prostaglandin 2. Further, PP13 causes stabilization of the surrounding connective tissues. Molecularly, this process requires the carbohydrate recognition domain to cross link between PP13 and molecules on the surface of the connective tissue and extracellular matrix.

### **4. Multiple Galectins and Deep Placentation**

In this article we focus on PP13, a member of the cluster of 6 galectins that emerged during primate evolution, and are only found in anthropoids. These species differ from their strep-sirrhine counterparts by having hemochorial placentas associated with a reduction in the number of offspring, with just one infant being common in monkeys, humans and apes. In all of these species, the newborns have relatively large brains and long gestations [112,113]. As described before, the success of pregnancy is mediated via increased blood flow to and from the placenta, which is achieved via an invasive hemochorial placentation [114–116]. The genetic differences between the mother and the fetal semi-allograft necessitate the development of immune tolerance to reduce the danger of fetal rejection by the mother, considering the alloantigen aspect of eutherian pregnancies [32,114]. We have provided evidence for the crucial role of galectin 13 (PP13) to render the mother immune-tolerant to sustain the hemochorial placentation during the long gestation of anthropoid primates. Interestingly, it has been pointed out that in addition to PP13, the other members of the cluster of galectins of chromosome 19 in anthropoids share high homology in their sequence and placental localization ([117] https://www.ncbi.nlm.nih.gov/kis/ortholog/29124/?scope=9526#genes-tab).

Table 1 indicates that humans have the entire cluster of which 5 are exclusively expressed in the placenta and one (Gal-10) is expressed in the bone marrow, but reaches the placenta via white blood cells, mainly eosinophils [2,31,32,39]. Orangutans, macaque, Sp. monkeys, and marmosets have four of the galectins, chimpanzees have two, and baboons, gorillas, and colobuses have only one galectin [31].



The presence of placental galectins in primate placenta is provided following the analysis of the evolutionary differentiation tree [31,32,36,39,115] with the exception of Galectin 10 (Gal 10) that is generated in bone-marrow but reaches the placenta via its expression in white blood cells. The letters A, B, and C reflect isoforms of the molecules. In terms of invasion: gorilla, chimpanzee, and human species have the deepest trophoblast invasion (3+ and 4+) reaching the inner myometrium [114–120]. The others have a much shallower implantation (1+ or 2+). The color code indicates in which species the Gal isoforms are expressed. The numbers in "total" refer to the numbers of species in which a specific Gal isoform is expressed. The numbers in "Count" refer to the numbers of Gal isoforms expressed in a given species.

Interestingly, all of the above species with the exception of the baboon, have PP13, most likely reflecting that this protein may be the first to be evolved or is derived from a common ancestor, and potentially it is the most essential one for a successful intrusive pregnancy [31]. The second most frequently found is Gal-14 that appears in four species. Gal-10 and Gal 16 are expressed in three species, while Gal-17 appears in two species and Gal-20 only in one. In terms of sequence homology, all galectins have close to 98% homology in the composition and configuration of their major amino acids of the carbohydrate recognition domain, the CRD. Gal-13 and Gal-16 share 73% amino acid sequence homology, while homology between Gal-13 and Gal-14 and between Gal-13 and Gal-10 is at the level of 68% and 57%, respectively. Interestingly, human Gal-14 and baboon Gal-14 have 98% amino acid sequence homology [34,36].

According to Carter et al. [118] there are different models of placentation among apes. Yet, gorilla, chimpanzee, and human species have the deepest trophoblast invasion and their remodeling of the spiral arteries occurs deep into their inner myometrium [114–120]. All three have Gal-13 (PP13), while baboons with a much shallower trophoblast invasion only express Gal-14. Thus, having multiple co-expression of galectins appears to be essential for successful invasive pregnancy, in which PP13 is pivotal but may not be the only one required. The interplay between the different galectins and their composition is now under study to understand their crucial role in normal pregnancy and pregnancy complications.

**Author Contributions:** M.S., H.M. and B.H. conceptualized the work; M.S., T.D. and M.M. performed the experiments used as the basis for this work; H.M., M.S. and S.G. provided the resources for this work; M.S. and H.M. did the original draft preparation; all authors reviewed and edited the text; B.H. and M.S. drew the figures; M.S., H.M. and S.G. have lead funding acquisition.

**Funding:** This research was funded by Daniel Turnberg Fellowship, UK Academy of Medical Sciences and the EU COST action CA16113 – CkiniMark to M.S. This study was also sponsored in part by the European Union (FP7) through the ASPRE project (601852) to H.M., S.G. and T.D. were sponsored by Hananja ehf, and Icelandic Research Fund (Rannís), gran<sup>t</sup> no. 163403-052.

**Conflicts of Interest:** The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Hamutal Meiri and Sveinbjorn Gizurarson have a patent for using PP13 for preeclampsia prevention. The other authors declare no conflict of interest.
