B lymphoproliferative diseases: Effective treatment, inhibited progression, and potential cures through isoform-specific targeting of the prolactin receptor

Abstract

In this chapter, we describe a potential new approach to treat lymphoproliferative diseases through isoform-specific knockdown of the long form of the prolactin receptor. The chapter includes a summary of the clinical and experimental links between prolactin and such diseases and presents sufficient background about prolactin and its receptors to explain the rationale for our approach. This background also aims to explain why clinical correlations between circulating prolactin and lymphoproliferative diseases may not appear as great as perhaps they are. In the final sections, we summarize our experimental evidence supporting the use of a splice-modulating oligomer that specifically targets the long form of the prolactin receptor. The work used mouse models of systemic lupus erythematosus and diffuse large B-cell lymphoma, human databases, and normal and malignant human cells. We also refer to previous and current studies using the splice-modulating oligomer which demonstrate its lack of toxicity, including in normal immune cells. For each section, we provide a take-home message in bold font so that the reader has the option to focus briefly or delve into details supporting the take-home message.

1. The clinical association between prolactin and B cell-mediated lymphoproliferative diseases

1.1. Autoimmune diseases

A functional immune system can distinguish self from non-self. If tolerance to self-tissues is broken, systemic inflammation and autoimmune diseases, including systemic lupus erythematosus (SLE), Sjogren’s Syndrome (SS), and rheumatoid arthritis (RA) result (Pisetsky, 2023). Among pathogenic immune cells in autoimmune diseases, B cells play a critical role. Dysregulated B cells and their terminally differentiated derivatives, plasma cells, produce self-recognizing autoantibodies that damage the skin, joints, heart, lungs, kidneys, and brain.

Of note, autoimmune diseases disproportionately affect more females than males; the female to male incidence ratios of SLE, SS, and RA are 9:1, 16:1, and 3:1, respectively (Brandt, Priori, Valesini, & Fairweather, 2015; Rider et al., 2018; van Vollenhoven, 2009; Weckerle & Niewold, 2011). Among women, autoimmune diseases peak during the reproductive years, although the disease can occur in children (Desai & Brinton, 2019; Izmirly et al., 2021). These epidemiological observations suggested an important role for female sex hormones in the pathophysiology. Estrogens have been the most thoroughly investigated, but our focus is on the hormone, prolactin, the production, and secretion of which is increased by estrogens (Grattan, 2015).

Increased circulating prolactin is associated with the exacerbation of symptoms of autoimmune diseases in both adults and children (De Bellis, Bizzarro, Pivonello, Lombardi, & Bellastella, 2005; El Miedany, Ahmed, Moustafa, & El Baddini, 2004; El-Garf, Salah, Shaarawy, Zaki, & Anwer, 1996; Fojtíková et al., 2010; Jara, Vera-Lastra, Miranda, Alcala, & Alvarez-Nemegyei, 2001; Jara et al. 2014, 2017; Ronis, Duffy, & Duffy, 2009; Soliman, Fahmy, Ali, & Shafie, 2023; Walker, Allen, Hoffman, & McMurray, 1995b; Yang, Li, Yang, & Li, 2016). Hyperprolactinemia in patients with autoimmune disease was found to positively correlate with increased levels of autoantibodies and abnormalities in the distribution of B-cell subsets (Liu et al., 2019). Further, inhibition of prolactin release with bromocriptine, which has no effect on estrogens, improves autoimmune disease symptomatology (Walker, 2001). Thus, there is clinical support for the premise that prolactin can drive autoimmune B-cell pathology.

1.2. B-cell malignancies

B-cell malignancies include mature lymphomas and lymphocytic leukemia (LL). In these diseases, transformed B cells evolve into overt malignancies through B-cell intrinsic and extrinsic mechanisms (Küppers, 2005; Terwilliger & Abdul-Hay, 2017). Unlike autoimmune diseases, the incidence of B-cell malignancies is similar in both sexes and literature supporting the existence of hyperprolactinemia in patients with B-cell neoplasms is limited to case studies and small patient cohorts (Foo & Sobah, 2014; Fridlyand et al., 2020; Grobe, Rühle, & Eckelmann, 1990; Kim et al., 2016; See, Cheuk, Fung, & Chan, 2013; Tovar-Méndez, Cuenca, Gómez-Pérez, & Cuevas-Ramos, 2022). However, autoimmune diseases with hyperprolactinemia have been widely shown to be associated with an increased risk of development of B-cell malignancies (Baecklund et al., 2003; Bernatsky et al., 2013, 2014; Fragkioudaki, Mavragani, & Moutsopoulos, 2016; Koff & Flowers, 2016; Morton et al., 2006; Simon, Thompson, Gandhi, Hochberg, & Suissa, 2015). Furthermore, prolactin promotes cycling and survival of B cells in culture (Berczi, Nagy, de Toledo, Matusik, & Friesen, 1991; Domínguez-Cáceres et al., 2004). Thus, prolactin may promote the evolution of B-cell malignancies.

2. What is important to know about prolactin in this context?

2.1. The molecule

Prolactin is a 23 kDa, four-helix bundle protein hormone/cytokine, produced in largest quantity by lactotroph cells of the anterior pituitary. In humans and mice/rats, the mature, secreted forms (i.e., minus the signal sequence) have 199 and 197 amino acids, respectively. There are three intramolecular disulfide bonds (formed between cysteines 4 and 11, 58 and 174, and 191 and 199 in the human) in the secreted hormone (Luck et al., 1992). These produce small N- and C-terminal, and a large central loop in the secreted molecule. In the secretory granules of the anterior pituitary lactotrophs, these cysteines in the sequence contribute to the formation of intermolecular disulfide bonds (Lorenson, 1985), the formation of which reduces the osmotic activity of, and therefore the energy required to maintain, the concentrated stored hormone. During exocytosis, these intermolecular disulfide bonds become intramolecular as proper folding of the monomers takes place (Greenan, Lorenson, Conconi, & Walker, 1990). The efficiency of this disulfide rearrangement decreases with increasing rapidity of prolactin secretion, making it more likely that multimers with reduced biological activity (Luck et al., 1992; Miyai, Ichihara, Kondo, & Mori, 1986) versus the monomeric hormone will be detected in the circulation under conditions resulting in hyperprolactinemia. Prolactin can also be posttranslationally modified, with N-linked glycosylation (Lewis, Singh, Sinha, & Vanderlaan, 1985) and phosphorylation (Chen et al., 1998) occurring in human prolactin, and O-linked glycosylation (Bollengier, Mahler, Braet, Claeyssens, & Vanhaelst, 2001) and phosphorylation (Oetting, Tuazon, Traugh, & Walker, 1986) occurring in rodent prolactin. Glycosylation of human prolactin changes its half-life in the circulation and importantly also its target tissue-specific biological activities (Hoffmann, Penel, & Ronin, 1993), while phosphorylation produces a form of prolactin that antagonizes the pro-proliferative and anti-apoptotic activities of unmodified prolactin (Chen et al., 1998; Ueda et al., 2006; Xu, Kreye, Kuo, & Walker, 2001). At the same time, the phosphorylated form can signal through the receptor to produce an enhanced state of differentiation (Huang, Ueda, Chen, & Walker, 2008; Schroeder, Brockman, Walker, & Schuler, 2003; Wu et al., 2021). Alternatively, with a high proportion of phosphorylated prolactin, the signal may result in apoptosis (Tan, Chen, Khoo, & Walker, 2011; Ueda et al., 2006a,b). In addition, prolactin can circulate covalently bound to IgG (Walker et al., 1995a) or non-covalently bound to other serum components (Kline & Clevenger, 2001; Lorenson, Ueda, Chen, & Walker, 2012). Prolactin can also be variously proteolytically cleaved (Clapp, Martial, Guzman, Rentier-Delure, & Weiner, 1993; Oetting, Ho, Greenan, & Walker, 1989), with N-terminal fragments possessing anti-angiogenic activity (Clapp et al., 1993). Thus, not all prolactin present in the circulation may be measurable by any one assay, and not all measurable prolactin may have a given biological activity. The existence of multiple forms of prolactin with diverse activities may reduce the degree of correlation between clinically or experimentally measurable serum prolactin levels and disease severity.

2.2. Regulation of pituitary prolactin secretion

The regulation of pituitary prolactin synthesis and release is complex and impacted by many factors (Ben-Jonathan, 2001). For the purposes of this chapter, the most important are stimulation of synthesis and release by 17β-estradiol (Grattan, 2015), increased release in response to both psychological and physical stressors (Faron-Górecka et al., 2023), and inhibition of synthesis and release by dopamine (Grattan, 2015). Thus, there are higher circulating levels in females versus males (Therkelsen et al., 2016), in individuals subject to acute or chronic stress (Faron-Górecka et al., 2023), and in patients administered drugs that decrease dopamine levels or function, such as classical antipsychotics (Faron-Górecka et al., 2023).

2.3. Biphasic dose-response curves

Because activation of the prolactin receptor requires binding of one prolactin molecule to two receptors (see below), very high levels of prolactin may produce a lesser response than lower levels. This is because with very high levels, it becomes more likely that inactive 1:1 (rather than 1:2) interactions between hormone and receptor will occur. Thus, very high levels of prolactin, such as may occur with prolactinomas, may not produce the same pathologies as levels closer to the normal range (e.g. Ewerman et al., 2020; Tworoger et al., 2013).

2.4. Heterologous prolactin

In animal model systems, we also now know that heterologous prolactin-receptor interactions (e.g. ovine prolactin in a rodent model) may not produce the same biological effect as the homologous hormone (Barbee, Radecki, Lorenson, & Walker, 2022). The response to heterologous hormone can be lesser than, non-existent, or opposite to the homologous hormone. Thus, there can be inconsistencies in effect observed by different investigators using the same animal model of disease but different species sources of prolactin.

2.5. Autocrine/paracrine prolactin

In addition to pituitary prolactin, many other tissues and cell types produce smaller quantities of prolactin that are used locally in an autocrine/paracrine manner, both as part of normal tissue physiology, as well as in pathological situations (Harvey, Martínez-Moreno, Luna, & Arámburo, 2015; Marano & Ben-Jonathan, 2014). The sequence of extrapituitary prolactin is identical to pituitary prolactin, although the regulation of expression is different (Marano & Ben-Jonathan, 2014). Very little is known about the regulation of prolactin expression in the B cell lineage. What we do know is that autocrine prolactin is overexpressed in B cell malignancies (Taghi Khani et al., 2023) and, like some other extrapituitary expression in humans, an additional untranslated exon (exon 1a) is present in the mRNA (DiMattia, Gellersen, Bohnet, & Friesen, 1988; Gerlo, Davis, Mager, & Kooijman, 2006; Reem, Ray, & Davis, 1999), which, at least in Jurkat T cells, confers responsivity to elevated cyclic AMP. The range of posttranslational modification of autocrine prolactin is as yet unknown, although there is one report showing that a proportion of autocrine prolactin can be glycosylated (Matera et al., 2000). What we do know is that, unlike pituitary prolactin, extrapituitary prolactin is not stored in concentrated secretory granules and its secretion is not inhibited by dopamine agonists (Ben Jonathan et al., 2008). Of particular importance to this chapter, multiple immune cell types, as well as stromal cells from lymphoid organs, have been shown to express prolactin in humans and mice (Marano & Ben-Jonathan, 2014; Montgomery, 2001; Wu, Devi, & Malarkey, 1996). However, expression by healthy B cells themselves seems to be limited to the mouse (Montgomery et al., 1987). Human naïve B cells express prolactin receptors but not prolactin. However, malignant human B cells express both prolactin and prolactin receptors, thereby creating a clear autocrine, in addition to potential paracrine, role for prolactin (Matera et al., 2000; Taghi Khani et al., 2023). Thus, pituitary-derived circulating prolactin is not the only relevant source of prolactin of importance in B cell-mediated disease. Therefore, treatment with dopamine agonists, which only affect release of prolactin from the pituitary, would be expected to be limited in their therapeutic efficacy, although this prediction is confounded by the presence of dopamine receptors on B cells (Morkawa, Oseko, & Morikawa, 1993). Since elevated autocrine prolactin may be insufficient to elevate circulating prolactin, normal levels of circulating prolactin do not rule out a role for prolactin in any given patient.

2.6. Prolactin and immunity

Pituitary influences on the immune system were recognized as early as 1905 when major reductions in the size of the thymus and spleen of rats occurred after hypophysectomy (Lundin, 1958 and earlier references therein; Nagareda & Kaplan, 1955). Specific interest in the role of prolactin, rather than other pituitary hormones, came from prolactin replacement protocols in hypophysectomized animals: prolactin replacement corrected deficiencies in both cell-mediated and humoral immunological functions (Nagy & Berczi, 1978; Nagy, Berczi, & Friesen, 1983), work that has subsequently been confirmed with recombinant prolactin to eliminate the possibility of contamination of older pituitary extract preparations with growth hormone (Richards & Murphy, 2000). In apparent contrast to the hypophysectomy experiments, prolactin or prolactin receptor knockout mice display no significant immune deficiencies (Bouchard, Ormandy, Di Santo, & Kelly, 1999; Horseman et al., 1997), suggesting cytokine redundancy and compensation for prolactin’s role when prolactin is absent throughout ontogeny; cytokine redundancy is a common phenomenon in the immune system (e.g. Riminton et al., 1998). However, when prolactin knockout mice are stressed, immune deficiencies become apparent (Dugan et al., 2002). Prolactin receptors have been reported to be present on many immune cells, including some T cells, B cells, NK cells, monocytes and macrophages, in addition to stromal cells in lymphoid organs (Orbach & Shoenfeld, 2007). Prolactin promotes T cell maturation and proliferation (Orbach & Shoenfeld, 2007), B cell survival and antibody secretion (Peeva & Zouali, 2005), enhances proliferation in response to antigen (Athreya, Pletcher, Zulian, Weiner, & Williams, 1993), and promotes production of pro-inflammatory cytokines (Orbach & Shoenfeld, 2007). Moreover, the diurnal rhythm of prolactin secretion inversely correlates with the expression level of CD25 on regulatory T cells (Bollinger, Bollinger, Naujoks, Lange, & Solbach, 2010). More specifically, prolactin inhibits the activation and therefore suppressive function of T regulatory cells (Legorreta-Haquet et al., 2012). T regulatory cells constitutively express the prolactin receptor and exhibit reduced activity (suppressive in this case) in the presence of prolactin (Legorreta-Haquet et al., 2012). Since T regulatory cell function suppresses autoimmune responses (Curotto de Lafaille & Lafaille, 2009), prolactin-mediated inhibition of T regulatory cells would be expected to promote autoimmune disease. Prolactin also increases expression of CD40L on T cells, and increases the expression of its receptor, CD40, found on antigen presenting cells, such as macrophages, dendritic cells and B cells (Biancone, Cantaluppi, & Camussi, 1999; Van Kooten & Banchereau, 2000). The interaction of CD40L and CD40 is required for B cell survival, activation, expansion, and differentiation, and hence antibody production (Law & Grewal, 2009). Constitutive activation of CD40 has been reported for lupus B cells (Zhang et al., 2012). Thus, the take home message from these few examples is that prolactin impacts immune responses by effects on multiple cell types.

3. What is important to know about the prolactin receptor?

3.1. Cytokine family membership

Prolactin receptors belong to the class I cytokine receptor family (Brooks, Dehkhoda, & Kragelund, 2018). They function as dimers and are single pass transmembrane receptors without intrinsic kinase activity (see below for signaling and Figs. 1 and 2 and associated legends for more detail). In the extracellular domain, they possess two fibronectin type III-like domains (S1 and S2) connected by a short linker region. The membrane distant fibronectin-like region, S1, is responsible for ligand binding, whereas the membrane proximal fibronectin-like region, S2, stabilizes the receptor off state such that constitutive signaling occurs in receptors lacking this region (ΔS2) (Huang, Tan, Chen, & Walker, 2015; Tan, Huang, Ueda, & Walker, 2008; Tan et al. 2013). Thus, prolactin functions through cytokine receptors.

Fig. 1. Prolactin receptors.

The figure illustrates pairs of long (LF, red) and a generic short (SF, blue) receptors with their sequence-identical extracellular domains (ECD), S1 and S2. The lipid interacting domains (LID) of the intracellular signaling regions are indicated by black arrows (Araya-Secchi et al., 2023). PM, plasma membrane. Prolactin is shown in green, interacting with the S1 region of the ECD, in such a way as to illustrate that it is not bilaterally symmetrical. As a result, binding to receptor 1 of the dimer is through different amino acids and has different binding affinity versus receptor 2 (Kossiakoff, 2004). Also shown is one constitutively associated signaling molecule, Jak (yellow), but other signaling molecules, as briefly described in the main text, may also be constitutively associated with one or more receptor type (Banks et al., 2024; Clevenger & Medaglia, 1994; Haxholm et al., 2015). Ligand binding induces a conformational change in the receptors and associated scaffolding and signaling molecules. For Jak, this results in transphosphorylation and activation. In the case of the long receptor, activated Jak2 can then phosphorylate tyrosines of the receptor, thereby allowing STATs (orange) or other signaling or scaffolding molecules to bind and become phosphorylated by Jak2 (depicted is the binding site for STAT5). Only the long receptor can directly activate STAT5. Jak is associated with all receptor isoforms and can, for example, activate the MAPK pathway via Shc, Grb2, Sos, Ras, Raf and MEK (purple diamonds). However, MAPK signaling through the short receptors is both delayed and prolonged versus the long receptor (Ueda et al., 2006b) and this difference is illustrated by the use of different purple arrows at the beginning of the cascade. Delayed and prolonged activation of MAPK can lead to differentiation and apoptosis (Lee, Jung, Chun, Kang, & Bang, 2002; Liu, Martindale, Gorospe, & Holbrook, 1996; Wu et al., 2005).

Fig. 2. Human and mouse receptors.

Some of the isoforms of the human and mouse receptors are illustrated more diagrammatically and arranged according to the length of their intracellular signaling domains. The numbering system reflects the order of discovery. The human intermediate form (IF) was named because of analogy to the previously discovered rat intermediate form. However, and confusingly, SF1a is longer. All receptors have sequence-identical extracellular and transmembrane (TMD) domains. Each of the receptors has a box 1 sequence (red), which is required for association with the signaling molecule, Jak, (Pezet, Buteau, Kelly, & Edery, 1997). A box 2 region (yellow), present in human LF, IF and SF1a and mouse LF may contribute to Jak2 binding/activation (Brooks, 2012).

3.2. Differential splicing

Multiple forms of the prolactin receptor are produced from a single gene product by differential splicing of pre-mRNA (Abramicheva & Smirnova, 2019; Pujianto, Curry, & Aitken, 2010). In humans, a long, intermediate, and three short forms have been detected, along with versions of the long and short forms missing one or other of the two extracellular domains, referred to as ΔS1 or ΔS2 (Figs. 1 and 2). To the authors’ knowledge, there has been no investigation of the existence of similar ΔS1 or ΔS2 versions of the intermediate form. Apart from when the S1 or S2 regions are missing, the extracellular domain of each of the receptors is identical. There are also reports of a soluble form of the receptor, which is essentially just the extracellular region (Kline & Clevenger, 2001), and additional isoforms missing exons (Kavarthapu & Dufau, 2022), but for these last, very little is known about their relative abundance, distribution, or function. In mice, there is a long form and 3 short forms (Fig. 2), and a fairly recent suggestion that there is a functional equivalent to the human intermediate form (Griffith et al., 2016). By contrast, only a long, intermediate, and one short form have been detected in the rat. Thus, the mouse alternative splicing shows greater analogy to the human situation, more likely making mice a better model for prolactin-mediated human disease. Different tissues/cells express different isoforms and/or ratios of the receptor isoforms, and expression of these isoforms is physiologically regulated (Abramicheva, Balakina, Morozov, Schelkunova, & Smirnova, 2019; Bridges, Scanlan, Lee, & Byrnes, 2011; Devi et al., 2009; Ouhtit, Morel, & Kelly, 1993; Ouhtit, Kelly, & Morel, 1994; Pujianto et al., 2010; Radecki et al., 2022; Schennink et al., 2015). For example, while the long form is predominant in many tissues, the short forms are expressed at higher levels in the human liver (Trott, Hovey, Koduri, & Vonderhaar, 2003) and predominate in the mouse liver (Barbee et al., 2022). In both mouse and human unstimulated B cells, the long form predominates, although there is a low level of expression of one short form (Taghi Khani et al., 2023). Thus, responses to prolactin will be different in different tissues/cells and under different physiological circumstances, and some cells may utilize constitutively active prolactin receptors, functioning without the necessity for ligand.

3.3. Signaling

To transduce a signal, prolactin receptors utilize several pathways, including Jak/STATs (most notably STATs 5, 3, and 1), PI3K/Akt, and MAPKs (Clevenger, Gadd, & Zheng, 2009). Additionally, the Src family kinases, including Fyn, Src, fak, Vav2 and ZAP-70 have also been linked to prolactin receptor signaling (Clevenger & Rui, 2022; Montgomery, Krumenacker, & Buckley, 1998). What combination of these pathways (and possibly others) is utilized will depend on the ratio of the different spliced forms of the receptor expressed, as well as the relative expression of different signaling molecules, receptor-associated scaffolding molecules, and terminators and inhibitors of signaling, such as suppressors of cytokine signaling (SOCS proteins) and protein inhibitors of activated STAT (PIAS proteins) in different cell types.

In humans, the long, intermediate and short forms with complete extracellular domains each use exons 3–9, which encode the extracellular and transmembrane domains, but incorporate various proportions of exon 10, ranging from the complete exon for the long form (LF), a fairly large proportion for the intermediate form (IF), a small proportion for the short form 1a (SF1a), to no part of exon 10 in short forms 1b (SF1b) and 1c (SF1c). Different lengths of exon 11 are added to the short forms (Ben-Jonathan, LaPensee, & LaPensee, 2008; Pujianto et al., 2010). That said, the SF1c has so far only been described in spermatozoa (Pujianto et al., 2010). In mice, the LF uses the same exons as the human, and for the short forms 1–3 (S1–3), exon 9 is linked to exons 11–13. Note that S1 uses exon 12, S2 uses exon 11, and S3 uses exon 13 (Banks, Trott, & Hovey, 2024; Tzeng & Linzer, 1997). These differences in 3′ pre-mRNA splicing create proteins with different intracellular signaling potential. For example, only the LF, with its translated complete exon 10 can signal through signal transducer and activator of transcription 5 (STAT5) (Ben-Jonathan et al., 2008), as illustrated in Fig. 1, left side. The SFs can act as dominant negatives for STAT5 signaling through the LF, but they also generate distinct signaling (Fig. 1, right side), leading to differentiated function and apoptosis (Devi et al., 2009, 2011; Huang et al., 2008; Tan & Walker, 2010; Trott et al., 2003). With pre-mRNA splicing dictating the outcome of signaling through the prolactin receptor, it becomes obvious that the relative expression of splicing factors will also be important to tissue/cell-specific functions of prolactin. Unfortunately, little is known about the splicing factors involved in the production of the different prolactin receptor isoforms. Somewhat more is known about alternative promoter usage in different tissues (Banks et al., 2024). Thus, tissue/cell-specific responses to prolactin (endocrine and/or autocrine/paracrine) depend on the relative expression of the different prolactin receptor isoforms.

3.4. Outcome of signaling through the prolactin receptor isoforms

Given both the regulation of total prolactin receptor expression and the different expression of signaling and receptor scaffolding molecules in different cell types, one cannot make absolute blanket statements about the role of each receptor type. In addition, the receptors can have different activities in the same cell when expressed individually versus in combinations (Le et al., 2012; Tan et al., 2008). Furthermore, differently posttranslationally modified prolactins can produce different signals through the same complement of receptors (Wu et al., 2003). However, for the purposes of this chapter, very simply put, work from multiple laboratories demonstrates that high relative expression of LF and/or IF receptors promotes cell proliferation and/or is antiapoptotic, whereas increased relative expression of the short receptors inhibits cell proliferation, is prodifferentiative and, when high enough, promotes apoptosis (e.g. Devi et al., 2009; Grible et al., 2021; Hartwell, Petrosky, Fox, Horseman, & Rogers, 2014; Huang & Walker, 2010; Lee, Walters, Reyland, & Anderson, 1999; Meng, Tsai-Morris, & Dufau, 2004; Schwertfeger et al., 2000; Tan & Walker, 2010; Tan et al., 2011; Ueda, Lo, Bartolini, & Walker, 2006b; Van Coppenolle et al., 2004; Wu, Ginsburg, Vonderhaar, & Walker, 2005; Wu et al. 2021). Thus, in many instances, cells that express high ratios of LF+IF: SF proliferate, and decreasing that ratio slows proliferation and promotes differentiation and/or apoptosis.

3.5. Knockdown of the LF/IF of the prolactin receptor

If prolactin stimulates cell proliferation through the LF/IF prolactin receptor, how then does one interfere with this process? While dopamine agonists that reduce pituitary prolactin release have been reported to ease autoimmune disease, they fail to eliminate autocrine and paracrine prolactin produced in specific tissues and hence have some, but limited, therapeutic value (Alvarez-Nemegyei et al., 1998). Although antibodies to the prolactin receptor (e.g. Novartis LFA102, (Minami, Ando, Tamura, Tajima, & Isaacs, 2020)) or pure prolactin receptor antagonists (Jomain et al., 2007) would interfere with both pituitary and autocrine prolactin, they again would be predicted to be less than optimal because they interfere with the function of all the receptor isoforms. Similarly, anti-prolactin antibodies, although somewhat effective at reducing the symptoms of SLE in a mouse model (Jiang et al., 2023), also interfere with both the disease-promoting and disease relieving forms of the prolactin receptor. As a reminder, the extracellular domain of both long and short isoforms is the same (see above). The outcome with antagonists and antibodies to the ligand or receptor, therefore, would be inhibition of both proliferative and differentiative/antiproliferative/apoptotic signals.

We therefore sought a way to knockdown only the LF/IF of the receptor so that prolactin would still be able to signal through the short receptors. In this way, we would remove the signaling to proliferation and leave intact differentiative, anti-proliferative, and pro-apoptotic signaling. As a result, we developed splice modulating oligomers (SMOs) that prevent the incorporation of exon 10 into the LF and IF in humans and the LF in mice (LFPRLR SMOs). Fig. 3 shows where the most efficient LFPRLR SMO hybridizes to human pre-mRNA. It is a 25 mer cDNA that sits ~half on the 3′ end of intron 9 and half on the 5′ end of exon 10. (Yonezawa et al., 2015). Although SF1a contains a small 154 nucleotide region of exon 10, we have yet to find any instance in which SF1a is knocked down with the LFPRLR SMO. We speculate that splicing factors prevent any effect on SF1a. Different splicing factor expression may also account for a lack of effect of LFPRLR SMO in T regulatory cells, which constitutively express LF prolactin receptors (Chen et al., 2021) and in naïve normal human B cells that express predominantly the LF (Taghi Khani et al., 2023). The LFPRLR SMOs efficiently penetrate cells/tissues because of their octaguanidine derivatization and are effective in mice at a dose of 100 pmoles/hour/mouse. No toxicities have been observed with up to 18 months of treatment (Chen et al., 2021; Taghi Khani et al., 2023; Yonezawa et al., 2015). Thus, we can examine the specific role of the LF/IF prolactin receptor in disease. In the process, we can gain valuable preclinical information about the potential use of the LFPRLR SMO as a therapeutic.

Fig. 3. Alternative splicing of human prolactin receptor pre-mRNA.

Exons in the pre-mRNA are shown as train carriages and introns as connectors in between. Translated exons include 3–9, with variations thereafter producing the differently spliced mRNAs. The portions of exon 11 incorporated into the human short forms are different but they are shown in this diagram just as part of the gray, exon 11 carriage. The LFPRLR SMO anti-sense DNA oligonucleotide hybridizes at the junction between intron 9 and exon 10 and prevents splicing of exon 9 to exon 10. The sequence of the human LFPRLR SMO is GCCCTTCTATTAAAACACAGACACA, which is different from the mouse version (Yonezawa et al.,2015). Although SF1a contains 154 nucleotides of exon 10, to date, we have not observed any effect of the LFPRLR SMO on expression of SF1a. We speculate that this is due to occupancy of the intron 9-exon10 hybridization site by a splicing regulator. In some cells (e.g. T regulatory cells, Chen et al., 2021) the LFPRLR SMO is not active, presumably also due to hybridization site occupancy. This technology is covered by U.S. patent 9909,128, held by the University of California, Riverside.

4. Long form prolactin receptor knockdown in systemic lupus erythematosus

4.1. Prolactin and autoimmunity

The link between hyperprolactinemia and the exacerbation of autoimmune symptoms in humans is mostly correlative. Fortunately, as noted earlier, mice appear to be an appropriate model in which to study prolactin-mediated disease. Of mouse models of autoimmune disease, the role of prolactin has been most extensively studied in SLE, likely because of the availability of models that resemble human disease. Administration of exogenous prolactin in SLE-prone mice increased autoantibody production and mortality from SLE, while inhibiting pituitary prolactin using dopamine agonists, reduced autoimmune symptoms (McMurray, Keisler, Kanuckel, Izui, & Walker, 1991; Peeva et al., 2003; Peeva, Gonzalez, Hicks, & Diamond, 2006). Analysis of signaling in B cells in response to increased circulating prolactin has shown increased phosphorylation of STAT proteins and enhanced survival of autoimmune B-cell subsets, leading to breakage of B-cell tolerance (Alemán-García et al., 2021; Carreón-Talavera et al., 2022; Flores-Fernández et al., 2016, 2021; Ledesma-Soto et al., 2012; Legorreta-Haquet et al., 2013; Peeva et al., 2003).

4.2. Knockdown of the LF receptor

All studies cited in the previous paragraph were conducted by administering exogenous prolactin and/or by inhibiting the production of pituitary prolactin in SLE-prone mice. Whether greater efficacy in autoimmune disease could be produced by inhibiting the effects of both pituitary and extra pituitary prolactin was unknown. Also unknown was whether overexpression of the LF prolactin receptor was pathogenic. Development of the LFPRLR SMO uniquely positioned us to fill these knowledge gaps. With the LFPRLR SMO, we systemically knocked down the LF receptor in vivo for 8 weeks before the onset of full-blown SLE in mice and relied solely on endogenous prolactin for receptor activation. We found the LF receptor to be overexpressed and the predominant isoform in B cells of SLE-prone mice. Knockdown of the LF reduced the numbers of B cells and their derivative plasma cell subsets. Both mature, long-lived plasma cells (LLPCs) and less mature plasmablasts were reduced. This demonstrated that signaling from the LF receptor modulates both the maintenance of existing plasma cells and the generation of new ones (Fig. 4). Importantly, knockdown of the LF receptor specifically rid the B-cell repertoire of cells with long complementary determining regions 3 (CDR3) of the variable portion of immunoglobulins; these cells are considered likely autoimmune B cells. Also deleted were B cells with non-productive immunoglobulin rearrangements, which are considered pre-malignant (Scheijen et al., 2019; Swaminathan et al., 2015). Because splenic B, T, and dendritic cells drive SLE pathology, we also examined the effect of LF knockdown on these cells. Both plasmacytoid and conventional dendritic cell numbers were reduced, whereas numbers of T cells and T-cell subsets (CD4⁺, CD8⁺, and pathologic CD4⁻CD8⁻ SLE T cells) remained unaltered (Taghi Khani et al., 2023). In normal mice, LF knockdown does not affect antigen-stimulated proliferation of splenic T cells or the ability of T regulatory cells to inhibit this response (Chen et al., 2021). Thus, knockdown of the LF prolactin receptor is an approach which seems capable of primarily targeting pathogenic B and dendritic cells in an autoimmune disease without compromising normal immunity.

Fig. 4. Response to increased prolactin leading to autoimmune disease.

LF prolactin receptors are shown in red and SF prolactin receptors in green. In response to elevated prolactin (yellow spheres), there is increased survival of immature bone marrow B cells as a result of their increased BCL2 expression. Unlike many hormones, prolactin upregulates its own receptor (Kavarthapu & Dufau, 2022), specifically LF receptors, which upregulation in turn increases the proliferative and anti-apoptotic effects of prolactin. B cells are disseminated to peripheral lymphoid organs. In the spleen, for example, they will interact with antigen-presenting dendritic cells, which are themselves increased in number by increased prolactin (Taghi Khani et al., 2023). Elevated prolactin increases expression of activation-induced cytidine deaminase (AID), which diversifies the B-cell repertoire through somatic hypermutation and class switch recombination and increases the frequencies of B cells with unproductive immunoglobulin heavy chain (IgH) rearrangements. Increased proliferation of B cells likely increases the chance of production and retention of abnormal plasma cells that secrete antibodies against self (Y, pink).

5. Long form prolactin receptor knockdown in B cell lymphoma

5.1. Disease initiation

B-cell malignancies evolve from a pre-malignant state where transformed B cell clone(s) acquire additional mutations and remodel their microenvironments to establish overt disease (Sungalee et al., 2014; Swaminathan et al., 2015; Tsai et al., 2008). That prolactin plays a critical role in the evolution of B-cell malignancies can be inferred from the following: (1) a sizeable number of patients suffering from B-cell lymphoma and leukemia exhibit hyperprolactinemia (Foo & Sobah, 2014; Fridlyand et al., 2020; Grobe et al., 1990; Kim et al., 2016; See et al., 2013; Tovar-Méndez et al., 2022). (2) in cell cultures, prolactin induces DNA synthesis in lymphoid cells and their cycling and survival through the induction of expression of oncoproteins such as c-MYC and BCL2 (Berczi et al., 1991; Domínguez-Cáceres et al., 2004; Kochendoerfer, Krishnan, Buckley, & Buckley, 2003). (3) human B-cell lymphomas express autocrine prolactin and prolactin receptors (Matera et al., 2000; Taghi Khani et al., 2023). (4) prolactin forms complexes with IgG that function via both prolactin and Fc receptors to induce proliferation of chronic lymphocytic leukemia (CLL) cells (Walker et al., 1995a). (5) autoimmune diseases, many of which have concomitant hyperprolactinemia, increase the risk of development of B-cell malignancies by >3-fold (Baecklund et al., 2003; Bernatsky et al., 2013, 2014; Fragkioudaki et al., 2016; Koff & Flowers, 2016; Morton et al., 2006; Simon et al., 2015; Zintzaras, Voulgarelis, & Moutsopoulos, 2005). (6) Estrogen, a stimulator of prolactin production, induces enzymes within autoimmune B cells (White et al., 2011) that are known to accelerate their transformation (Swaminathan et al., 2015; Tsai et al., 2008).

To determine if prolactin is an important contributor to malignant B cell evolution, we used the LFPRLR SMO to knock down the LF receptor in both SLE-prone mice, which have an increased risk of B-cell transformation, and in mice carrying pre-malignant B cells that make them prone to development of diffuse large B-cell lymphoma (DLBCL). In both models, knockdown showed that the prolactin-LF receptor axis was responsible for the increased B-cell numbers inherent to the disease. However, the mechanisms of malignant B-cell evolution downstream of prolactin varied. In SLE-prone mice, prolactin increased B-cell proliferation and the expression of the B-cell mutator enzyme, activation induced cytidine deaminase, thereby elevating the risk of B-cell transformation (Fig. 4). In DLBCL-prone mice, prolactin, working through the LF receptor, increased the expression of the BCL2 and TCL1 oncoproteins, thereby driving survival rather than proliferation or further mutation of pre-malignant B cells (Fig. 5) (Taghi Khani et al., 2023). Thus, by specifically signaling through the LF prolactin receptor, prolactin increases both the risk of B-cell transformation in SLE, and of establishment of DLBCL from pre-malignant B cells.

Fig. 5. Prolactin-induced progression and sustenance of B cell malignancies.

LF prolactin receptors are shown in red and SF prolactin receptors in green. In response to prolactin (yellow spheres), LF prolactin receptors are upregulated. Prolactin signaling through the LF receptor increases the expression of BCL2 and the oncogene, TCL1, increasing the survival of pre-malignant B cells and enhancing the risk of establishment of overt B cell malignancies. Malignant cells express autocrine prolactin (yellow spheres), enhancing local prolactin concentrations and signaling through the LF prolactin receptor. Signaling through the LF receptor induces the expression of MYC and BCL2 in overt human diffuse large B cell lymphoma and B-acute lymphoblastic leukemia (Taghi Khani et al., 2023), thereby promoting malignant B-cell growth and survival.

5.2. Effects on overt B-cell lymphomas

The importance of prolactin in patients with overt DLBCL is underscored by our observation that expression of high autocrine and paracrine prolactin in the lymphoma microenvironment predicts unfavorable clinical outcomes both at diagnosis and after standard therapies (Taghi Khani et al., 2023). This leaves open the possibility of dual therapy or second line therapy with the LFPRLR SMO. Expression of prolactin receptors and glycosylated and non-glycosylated forms of autocrine prolactin were discovered in human lymphoma cells two decades ago (Matera et al., 2000). We have now added to this initial observation by finding that normal human naïve B cells and human B-cell lymphomas differ markedly in their expression of autocrine prolactin and receptor isoforms: Normal B cells express the tumor-suppressive SF prolactin receptor and no autocrine prolactin. By contrast, their malignant counterparts exclusively express the tumor promoting LF prolactin receptor and autocrine prolactin (Fig. 5). Knockdown of the LF receptor in malignant human B cells reduces the phosphorylation of STAT3, lowers their expression of MYC and BCL2, and their survival. In vitro, autocrine prolactin is required for LF prolactin receptor-driven survival of malignant B cells (Taghi Khani et al., 2023). Use of the LFPRLR SMO therefore has the potential to eliminate transformed cells and to date has had no adverse effect on peripheral blood mononuclear cells derived from normal donors. Thus, the expression of autocrine prolactin combined with the exclusive expression of the LF receptor in established human B-cell lymphomas facilitates their growth and survival.

6. Potential limitations and additional benefits of treatment with the LFPRLR SMO

6.1. Potential limitations

1. If humans are like rodents, as they so far appear to be regarding expression of prolactin receptors, almost all tissues express some form of the prolactin receptor, although not all cells within any given tissue are positive (Aoki et al., 2019; Ben-Jonathan et al., 2008; Ouhtit et al., 1993, 1994; Tzeng & Linzer, 1997). Treatment with the LFPRLR SMO therefore has the potential to impact multiple tissues. Most affected, however, will be cells expressing high levels of LF receptor. In normal individuals, the most obvious example would be the lactating mammary gland (Ouhtit et al., 1993). Indeed, the major non mouse-specific phenotype of prolactin receptor null mice is the inability to lactate (Ormandy, Binart, & Kelly, 1997). We would predict therefore that the LFPRLR SMO would interfere with human lactation. By contrast, one would expect very little effect in cells predominantly expressing the SF receptor, such as hepatocytes (Ouhtit et al., 1994). Even with the broad distribution of prolactin receptors throughout the body, administration of LFPRLR SMO at anti-tumor doses (100 pmoles/hour/mouse) to normal mice resulted in no histologically detectable tissue abnormalities and no change in the rate of cell division in the small intestine. Furthermore, blood cell counts, behavior, and liver enzymes were all normal (Chen et al., 2021; Taghi Khani et al., 2023; Yonezawa et al., 2015).

2. So far, LFPRLR SMO has only been delivered via subcutaneous Alzet^™ minipump. From an experimental point of view, this has tremendous advantages, including a constant rate of release and no first-pass losses in the liver. Thus, clinical application will require testing of equivalent slow-release subcutaneous depot forms/mechanisms.

3. The LFPRLR SMO is so rapidly taken up by cells that it is difficult to perform traditional pharmacokinetics despite our development of a specific assay for its presence in biological fluids that is sensitive to levels below 1 pmol (Lorenson, Chen, & Walker, 2019). On the other hand, very efficient uptake by cells is the aim of drug delivery.

4. A concern with rapid uptake could be a gradation in effective delivery to tissues at a distance from the subcutaneous depot site, which we have seen no evidence for in mice (Chen et al., 2021; Taghi Khani et al., 2023; Yonezawa et al., 2015) but which could be a greater issue in human subjects, requiring different modes of delivery. However, for some applications, regional delivery may be advantageous.

6.2. Potential additional benefits

1. All analyses to date using normal human peripheral blood mononuclear cells show that their viability is unaffected by the LFPRLR SMO when treated ex vivo, even at high concentrations (>2.5 μM, Taghi Khani et al., 2023). This is a significant advantage over current therapies for B cell proliferative diseases, none of which discriminate between abnormal and normal cells, thereby leaving the patient in an immunocompromised state.

2. Overexpression of the LF prolactin receptor is associated with a growing number of proliferative diseases. These include, but are not limited to, prostate cancer (Dagvadorj et al., 2007), breast cancer (Meng et al., 2004; Yonezawa et al., 2015), and ovarian cancer (Sundaram et al., 2017; Tan et al., 2011). Development of the LFPRLR SMO for clinical use in B cell proliferative diseases could therefore lead to further clinical applications. Indeed, its efficacy in syngeneic and xenograft mouse models of breast cancer has already been demonstrated (Chen et al., 2021; Yonezawa et al., 2015).

3. The LFPRLR SMO is stable at room temperature for years, thereby reducing the cost of distribution and therefore treatment.

4. As a morpholino oligonucleotide, the LFPRLR SMO does not activate toll-like receptors and is resistant to nuclease degradation, both outside and inside cells.

5. The lack of toxicity observed to date, which extends up to 18 months of treatment of normal mice, suggests that the LFPRLR SMO could be used for long-term therapy of autoimmune disease and, in the process, decreasing the incidence of lymphoma development.

6. Clinical validation of the splice modulation approach may lead to the development of additional SMOs for a variety of diseases. This idea is discussed more generally for cancers in a recent publication (Chen & Walker, 2022).

7. Summary

1. Prolactin-LF prolactin receptor driven cell proliferation and/or survival is an important contributor to the lymphoproliferative diseases studied and, based on the literature, likely to others.

2. Knockdown of the LF prolactin receptor can have a therapeutic effect on lymphoproliferative diseases while not impacting normal immunity (additional information about normal immunity in Chen et al., 2021).

3. Because signaling through the tumor-suppressive SF prolactin receptor is preserved, treatment with the LFPRLR SMO is more effective than prolactin antagonists, or antibodies to the prolactin receptor or prolactin.

4. Because of the demonstrated lack of toxicity of the LFPRLR SMO in experimental animals and normal human immune cells ex vivo, the LFPRLR SMO has the potential to be an important therapeutic. It may even be a preventive therapy in instances where autoimmune disease predisposes to the development of malignancy.