Exploiting frequent and specific expression of PRL3 in pediatric solid tumors for first-in-child use of PRL3-zumab humanized antibody

Phosphatase of regenerating liver 3 (PRL3) is a specific tumor antigen overexpressed in a broad range of adult cancer types. However, its physiological expression in pediatric embryonal and mesenchymal tumors and its association with clinical outcomes in children is unknown. We sought to profile the expression of PRL3 in pediatric tumors in relation to survival outcomes, expression of angiogenesis markers, and G-protein-coupled receptor (GPCR)-mitogen-activated protein kinase (MAPK) signaling targets. PRL3-zumab, a first-in-class humanized antibody, was administered in a dose escalation schedule in a first-in-child clinical trial to study toxicity, pharmacokinetics, and clinical outcomes. Among 64 pediatric tumors, PRL3 was most frequently expressed in neuroblastoma (100%), rhabdomyosarcoma and non-rhabdomyosarcoma soft tissue sarcomas (71%), and renal sarcomas (60%) but absent in paired normal tissues. PRL3 was expressed in 75% of relapsed tumors and associated with shorter median event-free survival. Microarray profiling of PRL3-positive tumors showed elevation of angiogenin, TIMP1 and TIMP2, and GPCR-MAPK signaling proteins that commonly interacted with PRL3. The first use of PRL3-zumab in a pediatric patient saw no adverse events. A 28.6% reduction in maximum target lesion diameter was achieved when PRL3-zumab was administered concurrently with hypofractionated radiation. These findings support wider exploration of PRL3 expression in embryonal and mesenchymal tumors and further clinical application of PRL3-zumab in pediatric patients.


INTRODUCTION
Phosphatase of regenerating liver 3 (PRL3), encoded by PTP4A3, is a member of a class of dual-specificity protein tyrosine phosphatases consisting of three members: PRL-1, PRL-2, and PRL3. 1 PRL-3 is a C-terminal prenylated phosphatase induced in regenerating liver and was first found to be specifically overexpressed in colorectal cancer metastases. 2 Since then, PRL3 protein overexpression has been observed in a variety of cancers, [3][4][5] where it is generally associated with poorer prognosis and metastatic disease. 6,7][10][11] PRL3 has been reported to be an excellent oncotarget, with expression in 80.6% of tumors across a broad range of cancer types but not in any adjacent normal tissue. 10,11RL3 is undetectable in the major normal human organs. 10However, most studies have evaluated PRL3 expression in epithelial carcinomas more commonly found in adult patients, while the profile of PRL3 expression in pediatric cancers remains poorly understood.In Wilms tumor, the most common pediatric renal malignancy, PRL3 is associated with a 3.4-fold increase in risk of recurrence in favorable-histology disease. 12However, in other childhood embryonal cancers and mesenchymal tumors, PRL3 expression and its relationship to clinical outcomes, such as relapse, have not been described.
5][16] To date, the agent has completed initial Phase I first-in-human testing in adults, demonstrating a good safety profile with no grade 3 or 4 drug-related severe adverse effects (SAEs) observed 17 and has progressed to phase II clinical trials (ClinicalTrials.gov:NCT04452955 and NCT04118114; Drugtrials.org.cn:CTR20211180). 5These findings are consistent with in vivo safety assessments in murine and non-human primate models.This lack of toxicity in preclinical and clinical settings is due to the tumor-specific expression of PRL3 protein, which is largely absent in normal human adult tissues, 15 with PRL3 mRNA detected only in skeletal and cardiac muscle. 18,19In preclinical studies, PRL3 has been shown to play a limited role in specification of neural crest progenitors 20 and regulates cellular migration in embryonic kidneys without affecting cell viability. 21,22Therefore, we hypothesize that PRL3 may likewise be expressed in a highly specific manner in developmental cancers, facilitating safe application of PRL3-zumab in these diseases.However, PRL3 expression in the developing tissues of children and A B Figure 1.PRL3 is not expressed in a pediatric mouse normal tissue array (A and B) Full western blotting of PRL3 expression in various organs of a (A) 3-day-old and (B) 1-week-old male FVB mice.The breast and lung tumor tissue from a 14-weekold female FVB-MMTV-PyMT mouse were used as positive controls.High PRL3 expression can be detected at 20 kDa in the positive control.Kidney tissue of 1-weekold mouse expressed weak PRL3.Blots were probed with an anti-PRL3 monoclonal antibody.B-actin was used as loading control.
its toxicity profile in pediatric patients have not yet been described.In this study, we investigate the translational potential of PRL3 as a new therapeutic target in children to address the unmet clinical need for more targeted treatment of pediatric solid tumors.

PRL3 is overexpressed in a wide range of pediatric solid tumors and absent in normal tissues
While PRL3 oncoprotein is an established therapeutic tumor-specific oncotarget in mature adults, its baseline expression in developing tissues is not known.We have observed previously that PRL3 is undetectable in 13 major organs of adult FVB wild-type (WT) female mice but highly overexpressed in primary breast tumors and metastatic lung tumors of the spontaneous FVB/ MMTV-PyMT tumor model. 14To estimate PRL3 protein expression at human pediatricequivalent developmental time points, we performed full immunoblots on protein lysates of various organs from 3-day-and 1-week-old male FVB/MMTV-PyMT mice, given the known interspecies correlation of organ development-related transcriptomes. 23In contrast to control tumors from a 12-week-old female FVB/MMTV-PyMT mouse, PRL3 was largely absent in normal neonatal murine tissues, with only low expression seen in kidneys of the 1-week-old animal (Figure 1).This suggested that PRL3 was minimally expressed in developing organs in mice.
Next, we characterized PRL3 expression in pediatric solid tumors.Patients with newly diagnosed malignant solid tumors were recruited following informed consent, and excess tissue from initial tumor (primary), adjacent normal tissues, metastatic tumors, and relapse tumors was obtained and evaluated with full western blots (Table S1).In 64 specimens from 34 pediatric cancer patients, PRL3 was frequently expressed in pediatric malignant solid tumors and absent in matched normal tissues.Specifically, among embryonal tumors, PRL3 was present in neuroblastoma (7 of 7, 100%) and Wilms tumor (3 of 7, 43%) and absent in adjacent matched normal tissues (Figure 2A).Among pediatric sarcomas, PRL3 was present in osteosarcoma (4 of 8, 50%), rhabdomyosarcoma and non-rhabdomyosarcoma soft tissue sarcomas (NRSTSs) (5 of 7, 71%), and renal sarcomas of infancy (3 of 5, 60%) (Figure 2B).The data show that PRL3 protein was highly overexpressed in multiple pediatric solid tumors but not expressed in any of the adjacent normal tissues.
PRL3 is frequently expressed in pediatric solid tumors from patients with relapse and poor prognosis Because PRL3 was frequently expressed in a wide range of pediatric solid tumors, we sought to determine whether its expression was also related to clinical characteristics in these patients.We first studied matched pairs of tumor samples from biopsies taken at initial diagnosis and relapse (Table S2).In tumor pairs from a patient with neuroblastoma and a patient with rhabdomyosarcoma, PRL3 was absent at initial diagnosis but expressed in the relapsed specimen.
In two Wilms tumor pairs, PRL3 expression persisted from diagnosis to later relapse.Among three relapse osteosarcoma tumor pairs, PRL3 showed a mixed pattern of expression with detection at either time point in one pair and detection only at either initial diagnosis or relapse in the remaining two pairs.PRL3 expression could not be detected in the tumor pair from a patient with malignant mesothelioma.Overall, PRL3 was detected in 4 of 8 (50%) initial diagnostic biopsy samples and 6 of 8 (75%) relapsed tumor samples, suggesting that PRL3 overexpression is more frequently present in relapse samples than diagnostic samples (Figure 3A).
To understand the clinical characteristics associated with PRL3 expression, we compared PRL3 status among tumor types and in patients with metastatic versus localized disease.We also assessed the correlation between PRL3 status and event-free survival (EFS).Among 31 samples with complete clinical data available, PRL3 was A B Figure 2. PRL3 is overexpressed in a wide range of pediatric solid tumors and absent in normal tissues (A and B) Full western blot of PRL3 in pairs of primary tumor (T) and adjacent normal tissue (N) and metastatic tumors (M) from pediatric patients: (A) neuroblastoma and Wilms tumor and (B) osteosarcoma, rhabdomyosarcoma and non-rhabdomyosarcoma soft tissue sarcoma (NRSTS) and renal sarcoma.The DLD1 and HCT 116 human colon cancer cell lines were used as PRL3-negative and -positive controls (Ctrl).Blots were probed with an anti-PRL3 monoclonal antibody.GAPDH was used as a loading control.more frequently expressed in embryonal tumors (neuroblastoma and Wilms tumor) compared with sarcomas (osteosarcoma, rhabdomyosarcoma, NRSTSs, and renal sarcomas) as well as in patients with regional and distant metastases compared with localized disease (Figure 3B).In patients where survival data were available, EFS functions were estimated using the Kaplan-Meier method.The median estimated EFS duration was found to be shorter in patients with PRL3positive tumors (Figure 3C).Notably, this shorter median estimated EFS was observed regardless of whether PRL3 status was determined by the initial diagnostic biopsy or the latest specimen in cases with serial surgical excisions.The median estimated EFS for PRL3-positive patients was 3.9 months compared with 17.7 months for PRL3-negative patients based on PRL3 status at initial diagnosis, while the median estimated EFS for PRL3-positive patients was 14.6 months compared with 17.7 months for PRL3-negative patients based on PRL3 status of the most recent tumor biopsy.][26] Thus, we anticipate that high PRL3 expression in tumor biopsies of pediatric tumors could be predictive of a poor prognosis.
PRL3 expression in pediatric tumor specimens is associated with elevation of pro-angiogenic factors and phosphorylation of G-protein-coupled receptor (GPCR)-mitogen-activated protein kinase (MAPK) pathway members Because we observed PRL3 expression to be associated with poor prognostic features in a range of pediatric solid tumors, we next sought to understand which signaling pathways might be upregulated in PRL3-positive tumors, particularly angiogenic and GPCR-MAPK pathway targets that are known to be associated with PRL3 activation. 27,28We interrogated five cases of PRL3-positive tumors using a human angiogenesis antibody array of 43 associated protein targets.Expression of angiogenin, TIMP1, and TIMP2 was specifically elevated in the PRL3-positive tumor specimens but not in PRL3negative adjacent normal tissues (Figure 4A), with normalized net expression of these angiogenic factors increased in neuroblastoma, leiomyosarcoma, and rhabdomyosarcoma tumors (Figure S1).This observation was verified in corresponding full western blots using antibodies directed against these angiogenesis-related targets (Figure 4B).Consistent with these findings, PRL-3 is known to enhance tumor aggressiveness through regulation of matrix metalloproteinases, particularly via overexpression of angiogenin and invasionassociated enzymes of the MMP9-TIMP1/2 axis. 29,30cause PRL3 is also known to promote tyrosine and serine/threonine phosphorylation of diverse signaling proteins, particularly those of the GPCR-MAPK pathway, we sought to profile the phosphorylation status of these targets in relation to known PRL3 expression status.Tumor-normal pairs of representative cases of neuroblastoma (neuroblastoma 7T and 7N) and rhabdomyosarcoma (rhabdomyosarcoma 6T and 6N) were evaluated using a GPCR-MAPK microarray.Among 193 microarray targets, tumor expression of members of the cyclic AMP (cAMP)/protein kinase and oncogenic Ras/Raf protein families showed the greatest up-regulation in comparison with adjacent normal tissue (Figure 5A).Next, phosphorylation ratios of 84 proteins with non-phosphorylated and phosphorylated forms were determined and compared.In rhabdomyosarcoma, 30 proteins were phosphorylated and in neuroblastoma, 45 proteins were phosphorylated, with 16 proteins commonly phosphorylated in both (Figure 5B).To understand the association of these phosphoproteins with PRL3, a STRING protein-protein association network was constructed, comprising these commonly phosphorylated proteins along with PRL3.2][33][34][35] Our findings of frequent phosphorylation of GPCR-MAPK protein targets in pediatric tumors reflects data from adult cancers suggesting that PRL-3 may promote phosphorylation of multiple oncogenic proteins by functioning as a possible activator kinase. 27rst-in-child compassionate use trial of PRL3-zumab in combination with radiotherapy for a patient with recurrent metastatic rhabdomyosarcoma PRL3-zumab is a first-in-class humanized antibody drug that specifically targets PRL3-overexpressing tumors without damaging surrounding healthy tissues, targeting PRL3 externalized on the tumor cell surface. 14,15Given its established safety profile from the initial phase I trial 17 and ongoing phase II trials in adult patients with advanced solid tumors, PRL3-zumab was administered to a pediatric patient with multiply relapsed metastatic PAX7-FKHR fusion-positive rhabdomyosarcoma on a compassionate basis after diagnosis of a fifth relapse and disease progression.The prior treatment history from initial diagnosis to fourth relapse is described in Table S3.First, to establish baseline and current expression of PRL3, samples of the patient's primary tumor and two subsequent relapse tumors were assessed for expression of PRL3 by western blotting.PRL3 was not detected in the primary tumor, but high expression was detected in the first and second relapsed tumors (Figure 6A).Given these findings, and serial recurrence despite multiple lines of treatment, PRL3-zumab was added to the salvage treatment plan for the fifth relapse (Table S4).While awaiting commencement of PRL3-zumab treatment, the patient received two doses of bridging salvage chemotherapy with intravenous (i.v.) vinorelbine given 7 days apart.Subsequently, over 1 month, two sessions of hypofractionated palliative radiotherapy (RT; 2 Â 4 Gy) to the left chest tumors were given over two consecutive days each, prior to and concurrent with treat-ment with PRL3-zumab.Beginning at a dose range below the established recommended phase 2 dose for adults of 6 mg/kg every 2 weeks, PRL3-zumab was administered initially in incremental doses from 4 to 5.4 mg/kg/day, with the intention to increase gradually to the full adult dose (Table S4).As a precautionary measure for this first-inchild use of PRL3-zumab, the first dose was administered in two divided doses, while a shorter dosing interval of 8-10 days was used in view of the patient's advanced disease (Figure 6B; Tables S3-S5).

Safety and toxicity
Throughout the trial, the patient did not experience any inflammatory responses, adverse effects, or SAEs and tolerated the infusion very well even as PRL3-zumab was progressively increased to the full dose.The laboratory parameters remained stable after treatment with PRL3-zumab.Significant reduction of hemoglobin after chemotherapy and RT were restored to initial levels within a week.White blood cell (WBC) and neutrophil counts remained within the normal range.Absolute lymphocyte counts were low, which was likely due to prior chemotherapy and RT.Liver enzymes were within the normal range, apart from consistently elevated alkaline phosphatase (Table S6).

Quality of life
The patient weighed 14.2 kg at the start of the compassionate trial.
The Lansky scale was 70, and Eastern Cooperative Oncology Group (ECOG) performance status was 1.The patient displayed mild respiratory symptoms and required oxygen at night, but other organ functions were normal.A week after receiving the first dose, the patient remained clinically well, weighed 14.3 kg, was weaned off oxygen, no longer required morphine because pain had subsided, and could be nursed at home.Five days after the second dose of PRL3-zumab, the patient developed an intermittent left frontal headache but with no abnormalities on physical examination.The symptoms did not interfere with daily activities, and the patient was able to attend school for 3 days during that time and had also gained 0.4 kg.The patient was given paracetamol for relief and still did not require opiates or supplemental oxygen.

Efficacy
Following bridging chemotherapy and two doses of PRL3-zumab with hypofractionated RT, serum lactate dehydrogenase (LDH) levels decreased to a nadir of 212 U/L from the prior baseline of 315 U/L prior to the fifth relapse, suggesting clinical response to the multimodal therapy.Subsequently, LDH levels increased to a peak of 614 U/L after cessation of the trial, corresponding to the observed development of new progressive disease (Figure 6B, top panel).Serial chest computed tomography (CT) and whole-body PET/CT scans were taken for periodic efficacy assessment.Maximum target lesion (TL) diameter measured 4.9 cm at baseline prior to the fifth relapse and was reduced to 3.5 cm after initiation of the multimodal regimen incorporating PRL3-zumab-a reduction of 28.6% (Figure 6B, top panel).While the left chest TL showed improvement (Figure 6B, bottom panel), lesions in the right chest (non-TLs) showed progression, and new metastatic disease in the left eye orbit and liver were detected.
According to Response Evaluation Criteria in Solid Tumours (RECIST) 1.1 criteria, this response was categorized as stable disease (<30% reduction in maximum TL diameter).Thus, while the combination of RT followed by PRL3-zumab led to tumor reduction in the left chest, because of the onset of new distant disease, the salvage treatment was stopped in lieu of palliative care.

Pharmacokinetics
Pharmacokinetics (PK) samples were collected pre and post C1D1, C1D8, and C2D1 doses.Serum concentrations of PRL3-zumab at each time point were detected using ELISA, and non-compartmental model analysis of PK parameters for single-dose administration was performed.The patient C max was 71.23 mg/mL, area under curve (AUC) was 27,130.25 mg/mL h, and t 1/2 was 10.9 days (Figure 6C).The serum concentration and PK parameters of this pediatric patient were comparable with those of an adult patient administered a 6 mg/kg dose in the phase I clinical trial (Figure 6D).
In summary, in this first pediatric patient treated with PRL3-zumab on a compassionate dose escalation protocol, treatment was well tolerated with short-term improvement in quality of life and reduction of TL size.Unfortunately, because of development of distant metastatic disease, treatment had to be discontinued in pursuit of palliative care options.

DISCUSSION
In this study, we found PRL3 to be expressed in a wide variety of pediatric solid tumors at diagnosis and relapse but not in normal patient tissues and developing organs of neonatal mice.Pediatric patients with tumor expression of PRL3 more frequently displayed negative prognostic clinical characteristics; PRL3-expressing pediatric tumors also showed overexpression of angiogenic factors and cancer-associated kinases implicated in other PRL3-positive cancers.Together, the preclinical evidence indicated PRL3 to be a prevalent and specific cancer target in childhood tumors, which supported the compassionate first-in-child use of PRL3-zumab in a pediatric patient with multiply recurrent rhabdomyosarcoma.The dose escalation trial had no drug-related SAEs and yielded short-term responses in serum  markers and TL size.Though the treatment was initiated in the context of very advanced disease, PRL3-zumab demonstrated drug efficacy after the first dose in the pediatric patient with metastatic and recurrent rhabdomyosarcoma.The patient showed improved quality of life; he did not require oxygen supplementation or morphine, experienced weight gain, and was able to resume some daily activities, such as attending school for 3 days.It is likely that the drug efficacy was due to the combination of PRL3-zumab with RT.
Overall, PRL3-zumab stabilized disease in the pediatric patient while providing better quality of life, and with minimal side effects.These findings indicate the clinical potential of PRL3 as a therapeutic target and prognostic biomarker in pediatric solid tumors and the safety and efficacy of its corresponding antibody drug PRL3-zumab in this patient population.
7][38][39] Most notably, survival rates for pediatric brain and bone cancers and soft tissue sarcomas have plateaued in the last 2-3 decades, 36,40 with the least advancements seen in metastatic bone and soft tissue sarcomas. 41This may represent the limit of therapeutic gains from treatment intensification strategies, 42 considering also the acute burden of symptom distress and, in survivors of childhood cancer, late effects that substantially impair quality of life. 43,44These trends are concerning given the proliferation of pediatric precision oncology trials in recent decades, which, despite having facilitated increased identification of targeted therapeutic options, have also encountered low rates of treatment assignment because of the low frequency of actionable molecular alterations in childhood tumors. 45,46To date, kinase inhibitors targeting ALK-, TRK-, and ROS-rearranged tumors [46][47][48][49] and ALK-, RET-, and BRAFmutated tumors, [50][51][52] have shown the most success in pediatric brain and solid tumors with objective response rates of greater than 70%, but numbers of eligible patients are few.For example, NTRK fusions are found in only 3% of pediatric tumors 53 and ALK in 8%-12% of neuroblastoma. 54,557][58] Hence, there remains an urgent need for more effective but safe treatments for pediatric tumors, particularly sarcomas, relapse, and metastatic disease.
A major barrier limiting the development of cancer therapeutics is the intracellular location of most oncogenic driver proteins, which has traditionally precluded effective targeting of surface targets. 59In pediatric tumors, their low mutational burden further limits the number of altered surface proteins for neoantigen-based therapeutic strategies. 60Adopting an alternative strategy of targeting intracellular proteins with a humanized antibody could address this limitation.PRL3zumab has demonstrated preclinical efficacy against a wide variety of cancers expressing PRL3. 14,15,59,61,62Given the broad range of pediatric solid tumor types demonstrated in this study to also express this oncoprotein, this makes PRL3 an attractive novel therapeutic target for pediatric tumors.4][65] The safety data emerging from ongoing clinical trials with PRL3-zumab in adult patients has shown no drug-related SAEs (ClinicalTrials.gov:NCT03191682, NCT04452955, and NCT04118114; Drugtrials.org.cn:CTR20211 180). 5Our clinical experience from first-in-child use of PRL3-zumab also suggests that the agent should have a similarly favorable safety profile in children and potential to positively impact quality of life and tumor response.Therefore, further clinical studies in younger patients could be pursued to evaluate the utility of PRL3-zumab in pediatric tumor patients.
,27,67,68 This first study in pediatric tumors now extends the scope of diseases where PRL3 expression is negatively associated with clinical prognosis.
We also explored possible oncogenic mechanisms that might be responsible for the adverse behavior in childhood tumors and found that Src was commonly phosphorylated in neuroblastoma and rhabdomyosarcoma tumors.While this has also been described in other adult tumors expressing PRL3, [31][32][33][34][35] interestingly it was in human embryonic kidney cells where PRL3 was first observed to mediate Src activation via Csk down-regulation. 31Notably, the degree of Csk activity is positively correlated with ectodermal differentiation in embryonic stem cells and neurogenic differentiation in neuroblastoma, [69][70][71] though its role in regulating cellular differentiation in other developmental cancers is not known.These findings point to a need for further study of the mechanistic role of PRL3 signaling in embryonal tumors, particularly in relation to crosstalk with G-protein-coupled receptors and Src-family kinases.Another area that warrants future study is expression of PRL3 in mesenchymal tumors, particularly rhabdomyosarcoma and osteosarcoma, where there is currently a dearth of clinical and preclinical evidence.Also, given the rarity of pediatric cancers and the variety of different histological subtypes, larger profiling studies are required to validate our findings.
We found that PRL-3 was a cancer-specific marker frequently expressed in a wide variety of pediatric solid tumors, particularly neuroblastoma, osteosarcoma, and rhabdomyosarcoma.In the first clinical experience with PRL3-zumab in a pediatric patient, PRL3zumab again demonstrated a good safety profile and yielded short-term disease stabilization and improvement of quality of life when used in combination with hypofractionated RT.Further characterization of PRL3 expression in embryonal and mesenchymal tumors is needed, along with wider clinical experience with use of PRL3-zumab in children, possibly in the context of basket trials for pediatric cancers.

Patients and specimens
From 2015-2022, all pediatric patients with malignant solid tumors who were undergoing surgical biopsies or resections were prospectively recruited at KK Women's and Children's Hospital with institutional review board approval (protocols 2012/450 and 2014/2079, "Modeling, Analysis and Translational Therapeutics for Tumors of Childhood [MAT3CH]").Written consent was obtained from parents and assent from children.Where available, aliquots of excess tumor tissue from these routine surgical procedures were snap frozen for molecular analysis.Disease characteristics and survival outcome data were directly recorded or obtained from the Singapore Childhood Cancer Registry.Additional archival frozen tissues were obtained from the SingHealth Tissue Repository.

Animals
Animal study was performed in accordance with approved guidelines with Agency for Science, Technology, and Research (A*STAR) institutional animal care and use committee (IACUC) approval (IACUC 161130).Tissue arrays of 3-day-and 1-week-old FVB mice and 14-week-old Mouse Mammary Tumor Virus-Polyoma virus Middle T antigen (MMTV-PyMT) mice were collected after euthanasia.

Protein extraction and western blotting Protein extraction
Protein extraction and immunoblotting were performed as described previously.100 mg of tissue was suspended in 50 mL of radioimmunoprecipitation assay (RIPA) lysis buffer (Sigma) supplemented with a protease and phosphatase inhibitor cocktail (Roche) for 15 min at 4 C and homogenized by tissue homogenizer (Polytron).The supernatant of tissue lysates was collected after centrifugation at 13,000 Â g for 40 min at 4 C.For cultured cells, 5 Â 10 6 cells were lysed in lysis buffer and clarified as described above.Protein concentrations of tissue and cell lysates were estimated using a bicinchoninic assay kit (Pierce).After addition of 2Â Laemmli buffer containing dithiothreitol (DTT) (50 mM final concentration), samples were boiled and used immediately for western blotting or stored at À80 C until use.

Western blotting
Tissue and cell lysates were resolved on 14% SDS-polyacrylamide gels and transferred to nitrocellulose membranes before blocking and probing with the indicated primary antibodies at 1:2,000 dilution (PRL3) or 1:100,000 dilution (GAPDH) overnight at 4 C.After thorough washing with TBS-T buffer (20 mM Tris [pH 7.6], 140 mM NaCl, 0.2% Tween 20), the membranes were incubated with goat anti-mouse H+L IgG horseradish peroxidase (HRP)-conjugated secondary antibodies at a 1:5,000 dilution for 1 h, washed with TBS-T, and visualized using a chemiluminescent substrate (Millipore).

Angiogenesis array
Protein samples from tumor and matched normal tissues were used in the angiogenesis array.The angiogenesis array was performed according to the manufacturer's instructions (ab193655).In brief, each membrane was blocked with 1Â blocking buffer at room temperature for 30 min.After aspiration, 50 mg of protein samples (diluted in 1 mL of 1Â blocking buffer) was added onto the membranes and incubated overnight at 4 C.The following day, the membranes were washed three times with 1Â wash buffer I and incubated for 5 min each at room temperature.The membranes were further washed three times with 1Â wash buffer II for 5 min each.Subsequently, the membranes were incubated with 1Â biotinylated antibody cocktail for 2 h at room temperature.Then the membranes were washed three times with 1Â washing buffer I and three times with 1Â washing buffer II.1Â HRPconjugated streptavidin was added to the membranes, which were then incubated for 2 h at room temperature.Membranes were washed three times with 1Â wash buffer I and three times with 1Â washing buffer II.Finally, the membranes were visualized using the chemiluminescence substrate provided in the kit.The density of each spot was quantified using ImageJ (1.53C).The background signal was subtracted from the raw densitometry data, and the resultant data were normalized to the positive control.The expression levels of cytokines were compared between tumor and matched normal tissue samples.

GPCR microarray
Protein lysates obtained from selected frozen samples of tumor and matched normal tissues were applied to a GPCR-MAPK Pathway Phosphorylation Antibody Array (Full Moon BioSystems, USA) containing antibodies against 193 targets in 6 replicates printed on standard-size coated glass slides.Briefly, the arrays were first blocked with blocking solution (Full Moon BioSystems) for 30 min at room temperature and then incubated with the biotin-labeled cell lysates at 4 C overnight.After washing 3 times, the conjugated labeled proteins were detected using Cy3-conjugated streptavidin (S32355 streptavidin, Alexa Fluor 555 conjugate, Invitrogen).Hybridized array slides were scanned with a SureScan microarray scanner (G2600D, Agilent Technologies, Santa Clara, CA, USA) at 10 mm per pixel, and the single-channel image data were extracted and converted to Elist object files using Agilent Feature Extraction software.Read data were processed using limma on R (v.4.1.2) as follows.Normalization was performed using adaptive background correction, normalized between arrays, and averaged across duplicate spots.For quality control, raw intensities and read spot types were checked, and between-array MA plots (log intensity ratios vs. log intensity averages) and comparative density plots across arrays were created and inspected; also, sigma vs.A mean plots (residual variances vs. average log expression) were used to check the mean-variance relationship of the expression data.Fold change and standard error were estimated by fitting a linear model for each protein target, following averaging across duplicate spots, and then used to compute the phosphorylation ratio of phosphorylated and unphosphorylated protein pairs, as follows: phosphorylation ratio = (phospho tumor / unphospho tumor )/(phospho normal /unphospho normal ).Protein-protein interactions of commonly phosphorylated proteins and PRL3 in pediatric solid tumors were evaluated using STRING v.11.5. 72

First-in-child compassionate use trial
The first-in-child compassionate use trial of PRL3-zumab was conducted at KK Women's and Children's Hospital with written parental consent and child assent.In this custom protocol, 2 sessions each of hypofractionated palliative RT were given over 2 consecutive days before and during the PRL3-zumab treatment.The C1D1 dose of PRL3-zumab was commenced 10 days after the initial 2 RT sessions.
The drug was administered intravenously in 2 divided doses that were 6 h apart, each at 2 mg/kg, with a total of 56 mg; a full standard dose of PRL3-zumab is 6 mg/kg.The C1D8 dose of PRL3-zumab was administered as a single infusion at 4 mg/kg with a total of 57 mg.C2D1 was administered 10 days later at a dose of 4.8 mg/kg with a total of 70 mg, and C2D11 was administered at a dose of 5.4 mg/kg with a total of 80 mg.The primary endpoint was progression-free survival (PFS) according to RECIST 1.1 and iRECIST criteria. 73Secondary endpoints were overall survival (OS) and duration of response according to RECIST and iRECIST criteria.Safety was assessed by Common Terminology Criteria for Adverse Events (CTCAE) version 5 criteria with adverse event (AE), serious adverse event (SAE), vital signs, ECOG performance status (PS), physical examination, and clinical laboratory parameters.quality of life (QoL) was assessed based on general patient wellbeing and body weight.
PK analysis of PRL3-zumab in the pediatric clinical trial patient PK samples were collected pre and post dose of C1D1 (a.m. and p.m.), C1D8, and C2D1.ELISA was performed to detect the concentration of PRL3-zumab in serum.Briefly, 96-well plates coated overnight with glutathione S-transferase (GST)-PRL3 (1 ng) were blocked with 3% bovine serum albumin in PBS-0.05%Tween 20 prior to incubation with a serum sample (1:20,000) for 1 h at 37 C.After extensive washing, an HRP-conjugated anti-human antibody (Pierce) was added for 1 h at 37 C. Colorimetric development was performed using a Turbo-TMB substrate (Pierce) and stopped by acidification with 2 M H 2 SO 4 .Absorbance was measured at 450 nm using a plate reader (Tecan).Non-compartmental model analysis of PK parameters for single-dose administration was performed using PK Solver software.PK parameters for PRL3-zumab including C max , AUCinf, and t 1 / 2 were analyzed.

Figure 3 .
Figure 3. PRL3 is frequently expressed in pediatric solid tumors with relapse and poor prognostic features (A) Full western blot of PRL3 in initially diagnosed primary tumor (T) and relapsed tumor (R) tissue pairs from malignant pediatric tumors: neuroblastoma, Wilms tumor, osteosarcoma, mesothelioma, and rhabdomyosarcoma.DLD1 and HCT 116 cells were used as negative and positive controls (Ctrls), respectively.GAPDH was used as a loading Ctrl.(B) Comparison of PRL3 expression based on tumor type (sarcomas [osteosarcoma, rhabdomyosarcoma, NRSTS, and renal sarcomas] and embryonal tumors [neuroblastoma and Wilms tumor] (left panel, n = 34, p = 0.073, chi-square test) and based on the spread of tumor (right panel, n = 27, p = 0.411, chi-square test).(C) Eventfree survival (EFS) was compared using the Kaplan-Meier method according to PRL3 status at initial diagnosis (left panel) and according to PRL3 status of last/latest tumor biopsy (right panel).Shorter median estimated EFS was observed in patients with PRL3-positive tumors.

Figure 4 .
Figure 4. PRL3 expression is associated with elevation of the pro-angiogenic factors angiogenin, TIMP-1, and TIMP-2 in multiple pediatric solid tumors (A) Heatmap displaying unsupervised clustering of mean semiquantitative expression of pro-angiogenic factors detected on an antibody array applied to pediatric tumor samples and corresponding matched adjacent normal tissues.Expression of angiogenin, TIMP-1, and TIMP-2 was uniquely elevated in PRL3-positive tumor specimens but not in PRL3-negative normal adjacent tissues.(B) Verification of the observation in (A) by corresponding full western blots using directed antibodies against these proangiogenic targets to show angiogenin and TIMP-1 protein levels.

Figure 5 .
Figure 5. Microarray of expression of GPCR-associated proteins in pediatric solid tumors demonstrates phosphorylation of protein kinase and Ras/Raf proteins, which are associated with PRL3 (A) Scatterplot of mean normalized relative expression of GPCR microarray targets in tumor-normal pairs of representative cases of neuroblastoma (neuroblastoma 7T and 7N) and rhabdomyosarcoma (rhabdomyosarcoma 6T and 6N); members of the two protein families showing greatest up-regulation in tumor versus normal tissue are highlighted (protein kinase/cAMP, blue; Ras/Raf, red).(B) GPCR-associated proteins commonly phosphorylated in tumor samples of the representative rhabdomyosarcoma and neuroblastoma cases (phosphorylation is defined as phosphorylation ratio > 1).(C) STRING protein-protein association network of commonly phosphorylated proteins of representative pediatric solid tumors (from B) alongside PRL3, with known and predicted protein-protein interactions indicated by colored lines.Notably, PRL3 shares direct associations with members of the protein kinase and Ras/Raf families, which are commonly phosphorylated in pediatric rhabdomyosarcoma and neuroblastoma.

Figure 6 .
Figure 6.First-in-child compassionate clinical trial of PRL3-zumab in combination with chemotherapy and RT shows a reduction of the TL and a PK profile similar to adult data (A) Full western blot of PRL3 expression of biopsies of primary tumor at initial diagnosis (T) and first and second relapses (R).The DLD1 and HCT116 cell lines were used as negative and positive Ctrls of PRL3 protein expression, respectively; GAPDH is used as a loading Ctrl.(B) Multiple parameter plot showing serum LDH level and target lesion (TL) size over 21 months leading up to terminal demise.A blue line indicates the LDH level, orange bars indicate TL size, red circles indicate hypofractionated palliative RT sessions, and black arrows indicate the times when PRL3-zumab therapy was given.Encircled numerals indicate the representative contrast-enhanced axial CT images in the corresponding lower panel: (1) at conclusion of treatment of the fourth relapse, (2) at diagnosis of the fifth relapse, and (3) after conclusion of RT and PRL3-zumab therapy.Asterisks in CT images indicate TL nominated for iRECIST primary endpoint determination.(C) PK parameters calculated based on non-compartmental model single-dose therapy.(D) Serum concentration of PRL3-zumab in the first-in-child PRL3-zumab trial patient (top panel) compared with the serum concentration of PRL3-zumab in an adult patient (bottom panel).