We were initially interested in developing an assay that could be used in HTS to identify inhibitors of DPAP1, a protease that is expressed by the human malaria parasite Plasmodium falciparum
. This enzyme is an essential cysteine protease involved in the final stages of hemoglobin degradation 
. DPAP1 is refractory to genetic disruption 
and its inhibition blocks parasite growth both in vitro
and in vivo 
, thus making it a potentially valuable anti-malaria drug target. We also chose to focus on this target because proteases are generally difficult to express in their active form as most of them are translated as inactive pro-enzymes. Furthermore, expression of P. falciparum
proteins is especially challenging due to the A/T-rich nature of the genome of this organism. Additionally, DPAP1 is an ideal enzyme to demonstrate the value of our approach because we already identified a suitable ABP for this target 
, we had specific information about its substrate specificity 
, and we had access to recombinantly expressed enzyme for comparison of our method with standard screening assays 
DPAPs recognize the N-terminal amine of substrate proteins and are efficient at cleaving dipeptide fluorogenic substrates. Our SAR studies on DPAP1 indicated that proline at P2 (N-terminal amino acid) provides selectivity towards DPAP1 relative to other cysteine proteases in P. falciparum
such as DPAP3 or the falcipains 
. Additionally, the screen of a positional scanning library of 7-amino-4-methylcoumarin (AMC) substrates identified Arg as a preferred residue at P1 
. Therefore, we synthesized the (Pro-Arg)2
-Rho substrate to target DPAP1 in parasite lysates (). This substrate is cleaved twice by DPAP1 to produce free rhodamine.
Use of an ABP to identify a DPAP1-selective substrate in parasite lysates.
We chose a rhodamine-based substrate because free rhodamine emits at a wavelength (523 nm) that is high enough to be free from most of the auto-fluorescence background of compounds in a diverse library of small molecules. Using more traditional protease fluorogenic substrates, such as AMC-substrates for example, which emit at a lower wavelength, will likely result in an increase in the rate of false negatives during HTS, since a significant portion of the molecules in a library will likely emit light below 500 nm. Also, a bidentate substrate is likely to be more specific since it needs to be cleaved twice in order to produce an optimal signal.
To begin to assess the utility of this substrate for measuring DPAP1 activity in complex protein mixtures, we measured its apparent Km
value in parasite extracts (). Importantly, we obtained an apparent Km
value (36 µM) that was within experimental error of the value measured using recombinant DPAP1, suggesting that this substrate functions similarly against the native and recombinant enzymes. To test whether (Pro-Arg)2
-Rho is processed specifically by DPAP1, we treated parasite lysates for 1 h with increasing concentrations of FY01 - a fluorogenic ABP developed in our lab to target dipeptidyl aminopeptidases 
- and measured both DPAP1 labeling () and substrate turnover (). Because FY01 only labels DPAP1 in trophozoite lysates () it was straightforward to quantify labeling of the enzyme by SDS-PAGE. We found that the dose dependent labeling of DPAP1 by FY01 perfectly correlates with the extent of inhibition of (Pro-Arg)2
-Rho turnover () suggesting that the substrate is processed virtually exclusively by DPAP1.
Having demonstrated the selectivity of our substrate, we next wanted to determine if it could be used in a low-volume, 384-well plate assay. For the assay we continuously monitored substrate turnover for 5 minutes and then used the slope of the emission fluorescence over time as readout of enzyme activity. As a positive inhibition control, we used JCP410, a covalent inhibitor previously identified in our lab to be highly selective for DPAP1 
. Using this inhibitor and the continuous measurement of substrate processing, we could demonstrate that the assay has a signal-to-noise ratio (S/N) of 300 and an almost perfect Z’ factor (). Because end-point assays are generally preferred when screening large numbers of compounds, we adapted the assay into an end-point format by simply quenching the reaction after 10 min with acetic acid (data not shown). We evaluated this end-point assay in a fully automated format using a multi-channel peristaltic pump to dispense reagents into a 384-well plate, and a stackable plate reader. The end-point assay has a slightly lower S/N than the continuous assay, but the Z’-factor remains very high (0.9) (). Therefore, both the continuous and endpoint assays using the (Pro-Arg)2
-Rho substrate are highly sensitive assays that are suitable for use in HTS with a large numbers of compounds. Furthermore, the use the 384 well format allows screening of large libraries using a relatively small amount of parasite lysates.
Development of a DPAP1-specific HTS assay.
To demonstrate the general utility of our approach, we also applied the same methods to rat liver extracts. These samples are more complex than parasite lysates and contain a number of highly related papain fold cysteine proteases, in addition to the DPAP1 homolog Cat C. As expected, addition of FY01 to rat liver lysates resulted in labeling of multiple proteases targets (). However, only the labeling of rat Cat C (at 23 kDa) correlated with inhibition of (Pro-Arg)2-Rho turnover, suggesting that it is a highly selective substrate for Cat C in this system (). Furthermore, when the rat liver assay was converted to a 384-well plate format (20 µL reaction volume), we were also able to generate a high S/N and Z’ factor making it suitable for use in HTS (). Overall, these results demonstrate that our identified substrate can be used in multiple different extract systems and furthermore that ABPs that target multiple related enzymes can be used to identify specific substrates.
Cat C-specific fluorogenic assay in rat liver lysates.