Nitrocellulose-bound achromopeptidase for point-of-care nucleic acid tests

Enzymes are the cornerstone of modern biotechnology. Achromopeptidase (ACP) is a well-known enzyme that hydrolyzes a number of proteins, notably proteins on the surface of Gram-positive bacteria. It is therefore used for sample preparation in nucleic acid tests. However, ACP inhibits DNA amplification which makes its integration difficult. Heat is commonly used to inactivate ACP, but it can be challenging to integrate heating into point-of-care devices. Here, we use recombinase polymerase amplification (RPA) together with ACP, and show that when ACP is immobilized on nitrocellulose paper, it retains its enzymatic function and can easily and rapidly be activated using agitation. The nitrocellulose-bound ACP does, however, not leak into the solution, preventing the need for deactivation through heat or by other means. Nitrocellulose-bound ACP thus opens new possibilities for paper-based Point-of-Care (POC) devices.

Enzymes are catalytical proteins that constitute crucial tools in biotechnology [1][2][3] . One particular area that uses many enzymes is Nucleic Acid Amplification Tests (NAATs) 4 . These diagnostic tests are capable of target detection with high sensitivity and specificity 5 , using three steps: (i) sample preparation, (ii) DNA amplification, and (iii) DNA detection. Their main disadvantage however is the requirement of high-end equipment and highly trained personnel, for carrying out these steps, which limits their use in POC devices. To overcome these limitations, numerous NAAT systems have been developed for the POC [5][6][7][8] . Out of these efforts, paper-based microfluidic diagnostic systems, often termed µPADs 9 are showing high potential to minimize costs and enable mass-production of POC NAATs. Contrary to PCR, RPA is particularly well suited for integration into POC devices since it is an isothermal DNA amplification method, and therefore, does not require a thermocycler 6,10 .
Even though a number of techniques have been presented for paper-based amplification and detection 5 , much less attention has been devoted to the development of sample preparation, wherein cell lysis and nuclear acid purification must occur prior to amplification. The purification steps generally require the removal of all compounds present in the lysate, including the lysis reagents which may inhibit downstream processes such as DNA amplification and detection 5 .
To integrate sample preparation in µPAD NAATs, Whatman FTA™ paper, a proprietary material used to extract DNA from cells and preserve it at room temperature, has been used 11,12 . FTA™ paper, however, introduces amplification inhibitors and requires a series of washing steps which makes its integration into POC diagnostics difficult 11,[13][14][15] . This is an inherent limitation that occurs when utilizing chemicals for lysis, which denature proteins nonspecifically and cannot be deactivated.
Enzymes in solution can also inhibit or otherwise negatively affect downstream reactions, and therefore have to be deactivated. Peptidases form a subgroup of enzymes that catalyze the hydrolysis of proteins 16 . Peptidases such as proteinase K, papain and ACP are commonly used to digest tissues and cells [17][18][19] . ACP is a mixture of enzymes known to efficiently lyse Gram-positive bacteria and that has been used in POC systems [20][21][22] . In order to proceed with amplification or other downstream steps ACP must, however, first be deactivated. ACP deactivation is typically achieved by heat 5,[20][21][22] . The same is true in the cases of lambda exonuclease 6,8 and DNase I 23 . For integration into POC NAATs, this introduces complexity since the minimum temperature required for ACP deactivation is 80 °C 20 .
Therefore, there is a need to develop methods for NAAT sample preparation that allows the utilization of enzymes such as ACP, but omits the need for downstream heat deactivation or multiple washing steps. Here, we use nitrocellulose paper to immobilize ACP and enable its utilization preventing it from entering downstream solutions, eliminating the need to deactivate it. To test whether nitrocellulose-bound ACP activity is affected by mixing, we performed RPA following agitation of the RPA mix containing nitrocellulose-bound ACP (Fig. S3). We analyzed the amplification products from all experiments in the same electrophoresis gel to allow for quantification (Fig. S3). Agitation did not significantly change band intensity for the positive control (Fig. 3A). It did however, result in a highly significant reduction in the presence of nitrocellulose-bound ACP (Fig. 3A). In fact, the intensity for nitrocellulose-bound ACP band was not significantly different from that of the negative control (Fig. 3A). In contrast, the presence and agitation of plain nitrocellulose in RPA mixture did not inhibit RPA (Fig. S4A,B).
To assess whether ACP could be released from the paper during agitation, we utilized chip-based capillary electrophoresis to compare ACP in solution with the ACP content in water following the agitation of nitrocellulose-bound ACP (Figs. 3B, S5). No release of ACP (30 to 50 kDa) was detected following agitation (Fig. 3B).
Furthermore, we investigated the activity of immobilized ACP on a reaction under passive diffusion by incubating nitrocellulose-bound ACP in RPA mix for 60 min (undisturbed) before initiating the RPA reaction (Figs. 3C, S6). The significant inhibition of RPA (Fig. 3C) can be plausibly explained by the diffusion of active RPA reagents to the surface of nitrocellulose where they reacted with the immobilized, yet active, ACP.
To assess the stability of nitrocellulose-ACP bond, we placed the nitrocellulose paper in water for 60 min, after which we analyzed it for traces of ACP by chip-based capillary electrophoresis (Figs. 3D, S7). Similarly, to the results shown in Fig. 3B, we did not detect any ACP with this method (Fig. 3D).
Nitrocellulose-bound ACP appeared stable when examined by chip-based capillary electrophoresis (Fig. 3B,D). The absence of bands from these samples compared to free ACP in solution suggests that ACP does not easily detach from the nitrocellulose (Fig. 3B,D). The same conclusion is supported by the fact that water, in which nitrocellulose-bound ACP had been thoroughly agitated, failed to affect the RPA reaction (Fig. 2B).
It seems that ACP can be stored in nitrocellulose without significantly affecting its function (Fig. 3A,C) and importantly, it can be used while immobilized (Fig. 3A,C), which allows for its removal from solution (Figs. 2B, 3B,D) without the need for heat or other deactivation. This is of particular relevance for the integration of enzymatic systems in portable devices where it is advantageous to have protocols with few and simple steps without external instrumentation 1,34 .

Conclusion
We demonstrated that nitrocellulose-bound ACP does not immediately inhibit RPA amplification in a stationary solution, contrary to ACP in solution. Nitrocellulose-bound ACP can, however, be activated by agitation in solution without being released. The described mechanism opens the possibility to utilize nitrocellulose-bound ACP for reactions such as cell lysis in paper-based NAATs, or possibly for reaction inhibition in other applications. Once the bound ACP has carried out its function it can easily be removed from the solution without the need for instruments and without using heat deactivation or other processes that might affect downstream applications. We believe that this article paves the way for the improvement of POC devices by facilitating the integration of e.g., an instrument-free lysis step.  Fig. S2. Agitation of nitrocellulose-bound ACP in water did not release enough ACP to significantly inhibit RPA (after the removal of the nitrocellulose, the water was used in the RPA). Furthermore, water in which plain nitrocellulose was agitated, does not significantly inhibit RPA either (n = 5 for all conditions, unpaired t-test, mean with SD). To examine the effect of agitation, nitrocellulose containing ACP was agitated in 10 μl nuclease-free water by pipetting 50 times and stirred thoroughly using the pipette tip. This water was applied directly to the RPA mix.  Fig. S3 showing that nitrocellulose-bound ACP is capable of RPA inhibition when mixed (n = 5 for all conditions, unpaired t-test, mean with SD). (B) No ACP (30 to 50 kDa) was detected by chip-based electrophoresis in water where nitrocellulouse-bound ACP was submitted to agitation. Full-length gel image is presented in Fig. S5. (C) Densitometric analysis of agarose gel electrophoresis results in Fig. S6 showing that nitrocellulose-bound ACP is capable of high to complete RPA inhibition within 60 min, possibly by diffusion of amplification enzymes to the ACP on the paper surface (n = 5 for all conditions, unpaired t-test, mean with SD). (D) No ACP (30 to 50 kDa) was detected by chip-based electrophoresis in water where nitrocellulose-bound ACP was left for 60 min. Full-length gel image is shown in Fig. S7.