Human Kallikrein Targeted Radioimmunotheranostics for Prostate Cancer Treatment "Presentation" - Hans David Ulmert
April 29, 2025
At the 2025 UCSF-UCLA PSMA Conference, Hans David Ulmert discusses targeting human kallikreins (KLK2/hK2 and KLK3/PSA) with radioimmunotheranostic approaches for prostate cancer. He explains these proteins are highly specific to prostate tissue, AR-driven, and normally confined within the gland until pathology causes leakage. Dr. Ulmert describes antibodies targeting the catalytic cleft inaccessible in circulating forms.
Biography:
Hans David S. Ulmert, MD, PhD, Lab Head, Associate Professor, Department of Molecular and Medical Pharmacology, University of California, Los Angeles (UCLA), Los Angeles, CA
Biography:
Hans David S. Ulmert, MD, PhD, Lab Head, Associate Professor, Department of Molecular and Medical Pharmacology, University of California, Los Angeles (UCLA), Los Angeles, CA
Read the Full Video Transcript
Hans David Ulmert: Thank you so much for inviting me to give a talk on hK2 and PSA-targeted radioimmunotheranostic for prostate cancer. Human prostate kallikreins-- they are KLK2 and KLK3-- they are very specific for human-- they're only found in humans, dogs, and old-world monkeys. And we are also the only species that spontaneously develop benign prostate hyperplasia and prostate cancer. They are specifically expressed at very abundant levels in prostate tissues, both malignantly derived and healthy prostate tissues. They are specifically driven by the AR, androgen receptor, meaning AR activity goes up, KLK2 and KLK3 goes up.
KLK2 and KLK3 are very similar. They were actually a duplication, phylogenetically, of each other. They have an 80% amino acid sequence identity. Still, they have a different substrate specificity. They are normally in a healthy prostate gland, kept very confined within the gland. However, when there is some type of pathology, if it's prostatitis or if it's benign prostate hyperplasia or prostate cancer, we see that small amounts of this are leaked out. So in terms of how much is leaked out is every millionth molecule. So it's not much compared to what is actually found in the tissues.
So if we look here at the expression in different tissues and we compare it with PSMA, for example, we see that it's very, very specific to prostate tissues and prostate cancer adenocarcinoma. We have made two antibodies, one for hK2 and one for PSA, and we're targeting the catalytic cleft. And that's important because when KLK2 and KLK3 are released into the blood circulation, they are immediately inhibited by protease inhibitors. So we're targeting the catalytic cleft, which is not found in the blood circulation.
And the endocytosis of this is driven by FcRn. Not to spend too much time on this, but, basically, it's been shown by Bloomberg's lab in 2005 that FcRn is normally prolonged in the blood circulation. However, if the antibody is bound to a protein, it's not spit out again into the blood circulation, but it's then endocytosed by the cell and is broken down, and that's how the antibody is then ending up in the cell. And you can see here, if we are ablating the ability of the antibody to complex with FcRn, we don't see any uptake at all.
Here is an example of Zirconium-labeled hK2 antibody. In a LNCaP model, we see that we have a higher uptake in VCaPs, which are expressing higher amounts. And we have a slightly lower than in LNCaPs, which are expressing less than VCaP.
We have tested this in orthotopic models, both osteoblastic and osteolytic combined with osteoblastic. And you can see here, we also use the LNCaP-AR model to see if we can then measure the AR output or the AR activity. And what we can see here is that we have a zirconium image in the leg of the mouse before castration. Then we castrate the mouse. And then 10 days later, we imaged him again. And what we see is that the hK2 signal, when it's quantified, correlates with the AR-driven output.
We also tested this in transgenic mice. In the left part of the section, you can see a normal mouse, and then a KLK2-expressing mouse in the middle that has a healthy prostate. And then that one is crossed with the Hi-MYC model. So we have spontaneous development of the inner carcinoma. And as you can see, it's taken up.
We can also see that we have uptake in GR-driven AR-resistant or enzalutamide-resistant tumors. Here is an example then in the LREX model, where we have developments of tumors in the liver. We also see that when we label it with zirconium, compare we label it with actinium, that in both the GEMM models and in LNCaP and VCaP models, that they correlate.
If we then look at the actinium-labeled antibody, KLK2 antibody, and we test that in Hi-MYC models that are expressing KLK2, we can see then that we have a very robust and highly functional therapeutic effect here over time.
Why the targeting of KLK2 and KLK3 is very positive is that given that we have a DNA repair, DNA repair is in upregulating the AR and AR is in upregulating the target, meaning that the smaller-- the more you hit the tumor, the smaller the tumor becomes, but the more antigen you are expressing.
Mike Morris has translated this, the hK2 antibody, into patients. First, they evaluated uptake with an indium-labeled version of the antibody. And in short, they saw that if they found uptake of the antibody in lesions, that could be detected with other modalities, including PSMA PET.
These are results from the actinium trial, also given to me by Mike Morris. And this is a year ago. So hopefully, we have even more results. But as you can see here, they are responding quite well. We also see that either they respond very well with robust and deep response or they're not responding at all. However, we can also see that responders can be hit a second time with a second dose and you see a response.
So in summary, the hK2 antibody can be utilized for both PET, SPECT, and for alpha and beta therapy. It's been translated to patients and has also been utilized as a bispecific and as a CAR-T by Janssen. If we then look at the PSA-targeting antibody, we've labeled that with both-- here, our example, with yttrium, and actinium, and zirconium.
And we've also tested it then and looked at the uptake in GEMM models that are specifically expressing PSA, and that we see that we have an uptake that correlates to where we have signal in the different prostate lobes, and that we see a similar uptake when we compare yttrium, actinium, and zirconium. When we test in subcutaneous tumors, we can see then that we have a faster effect with yttrium compared to actinium, but actinium has then a longer, more deep and durable effect.
We validate this in monkeys. Monkeys have a 500 to 1,000-fold lower PSA expression compared to what we have in humans. Despite this, we can then see-- oops. Despite this, we can then see uptake specifically in the prostate despite that we have very, very low uptake in PSA expression in monkeys.
So in summary, for the 5A10 antibody targeting PSA, it's been licensed to Radiopharm Theranostics. And from what I gather, is that it's now GMP produced. And hopefully, it will go into the clinic soon. The upside of using KLK2 compared to KLK3 is that you have higher avidity of KLK3 compared to KLK2. And that was it. Thank you very much.
Hans David Ulmert: Thank you so much for inviting me to give a talk on hK2 and PSA-targeted radioimmunotheranostic for prostate cancer. Human prostate kallikreins-- they are KLK2 and KLK3-- they are very specific for human-- they're only found in humans, dogs, and old-world monkeys. And we are also the only species that spontaneously develop benign prostate hyperplasia and prostate cancer. They are specifically expressed at very abundant levels in prostate tissues, both malignantly derived and healthy prostate tissues. They are specifically driven by the AR, androgen receptor, meaning AR activity goes up, KLK2 and KLK3 goes up.
KLK2 and KLK3 are very similar. They were actually a duplication, phylogenetically, of each other. They have an 80% amino acid sequence identity. Still, they have a different substrate specificity. They are normally in a healthy prostate gland, kept very confined within the gland. However, when there is some type of pathology, if it's prostatitis or if it's benign prostate hyperplasia or prostate cancer, we see that small amounts of this are leaked out. So in terms of how much is leaked out is every millionth molecule. So it's not much compared to what is actually found in the tissues.
So if we look here at the expression in different tissues and we compare it with PSMA, for example, we see that it's very, very specific to prostate tissues and prostate cancer adenocarcinoma. We have made two antibodies, one for hK2 and one for PSA, and we're targeting the catalytic cleft. And that's important because when KLK2 and KLK3 are released into the blood circulation, they are immediately inhibited by protease inhibitors. So we're targeting the catalytic cleft, which is not found in the blood circulation.
And the endocytosis of this is driven by FcRn. Not to spend too much time on this, but, basically, it's been shown by Bloomberg's lab in 2005 that FcRn is normally prolonged in the blood circulation. However, if the antibody is bound to a protein, it's not spit out again into the blood circulation, but it's then endocytosed by the cell and is broken down, and that's how the antibody is then ending up in the cell. And you can see here, if we are ablating the ability of the antibody to complex with FcRn, we don't see any uptake at all.
Here is an example of Zirconium-labeled hK2 antibody. In a LNCaP model, we see that we have a higher uptake in VCaPs, which are expressing higher amounts. And we have a slightly lower than in LNCaPs, which are expressing less than VCaP.
We have tested this in orthotopic models, both osteoblastic and osteolytic combined with osteoblastic. And you can see here, we also use the LNCaP-AR model to see if we can then measure the AR output or the AR activity. And what we can see here is that we have a zirconium image in the leg of the mouse before castration. Then we castrate the mouse. And then 10 days later, we imaged him again. And what we see is that the hK2 signal, when it's quantified, correlates with the AR-driven output.
We also tested this in transgenic mice. In the left part of the section, you can see a normal mouse, and then a KLK2-expressing mouse in the middle that has a healthy prostate. And then that one is crossed with the Hi-MYC model. So we have spontaneous development of the inner carcinoma. And as you can see, it's taken up.
We can also see that we have uptake in GR-driven AR-resistant or enzalutamide-resistant tumors. Here is an example then in the LREX model, where we have developments of tumors in the liver. We also see that when we label it with zirconium, compare we label it with actinium, that in both the GEMM models and in LNCaP and VCaP models, that they correlate.
If we then look at the actinium-labeled antibody, KLK2 antibody, and we test that in Hi-MYC models that are expressing KLK2, we can see then that we have a very robust and highly functional therapeutic effect here over time.
Why the targeting of KLK2 and KLK3 is very positive is that given that we have a DNA repair, DNA repair is in upregulating the AR and AR is in upregulating the target, meaning that the smaller-- the more you hit the tumor, the smaller the tumor becomes, but the more antigen you are expressing.
Mike Morris has translated this, the hK2 antibody, into patients. First, they evaluated uptake with an indium-labeled version of the antibody. And in short, they saw that if they found uptake of the antibody in lesions, that could be detected with other modalities, including PSMA PET.
These are results from the actinium trial, also given to me by Mike Morris. And this is a year ago. So hopefully, we have even more results. But as you can see here, they are responding quite well. We also see that either they respond very well with robust and deep response or they're not responding at all. However, we can also see that responders can be hit a second time with a second dose and you see a response.
So in summary, the hK2 antibody can be utilized for both PET, SPECT, and for alpha and beta therapy. It's been translated to patients and has also been utilized as a bispecific and as a CAR-T by Janssen. If we then look at the PSA-targeting antibody, we've labeled that with both-- here, our example, with yttrium, and actinium, and zirconium.
And we've also tested it then and looked at the uptake in GEMM models that are specifically expressing PSA, and that we see that we have an uptake that correlates to where we have signal in the different prostate lobes, and that we see a similar uptake when we compare yttrium, actinium, and zirconium. When we test in subcutaneous tumors, we can see then that we have a faster effect with yttrium compared to actinium, but actinium has then a longer, more deep and durable effect.
We validate this in monkeys. Monkeys have a 500 to 1,000-fold lower PSA expression compared to what we have in humans. Despite this, we can then see-- oops. Despite this, we can then see uptake specifically in the prostate despite that we have very, very low uptake in PSA expression in monkeys.
So in summary, for the 5A10 antibody targeting PSA, it's been licensed to Radiopharm Theranostics. And from what I gather, is that it's now GMP produced. And hopefully, it will go into the clinic soon. The upside of using KLK2 compared to KLK3 is that you have higher avidity of KLK3 compared to KLK2. And that was it. Thank you very much.