3. Applications of NAPPA Technology
Next, different applications of NAPPA technology are described showing several studies and their results. These applications are classified according to protein-protein interactions studies, vaccine development and the evaluation of autoimmune responses (Figure 3). Table 1 summarizes the main studies developed in the field and described in the text.
Figure 3  Applications of NAPPA technology.

3.1. NAPPA Technology for Understanding Proteins-Proteins Interactions
Four years after the first design of NAPPA technology, LaBaer’s group confirmed that protein function is maintained in printed proteins on high-density arrays. With this purpose, they designed an array expressing 647 unique genes in duplicate and tested for several well-characterized interactions, Jun-Fos and p53-MDM2 among others. Simultaneously, they expressed the corresponding protein printed on the array and co-expressed the query protein by adding the appropriate cDNA to the cell-free expression lysate. Using specific antibodies against Jun, Fos and MDM2, they detected specific interactions of these proteins. It is also necessary to take into account that protein function can be compromised by lack of PTMs and/or misfolding of certain domains due to the absence of chaperones and cofactors. Concerning the lack of PTMs, it is possible to use alternative cell free expression systems depending on the protein to be expressed. Thus, different expression systems have been developed (including HeLa, Leishmania, E. coli, rabbit expression systems, among others). Also, including ribosomal machinery and chaperones (such as HSP90 or HSC70) may encourage the folding of large multi-domain proteins [9].
More recently, in 2012, Fuentes and collaborators published a work in which a total of 450 mRNAs from O. moubata tick salivary glands were extracted and purified, and then transfected into a donor vector (pDONR222) generating a library of cDNA. Finally, this library was transfected again into a library destination expression vector (pANT7_GST), which allows in situ expression of GST-tagged proteins in cell-free systems. They built a NAPPA array randomly choosing 480 clones with validated sequences. After confirming successful display of the recombinant fused GST tag protein, the correct display of individual tick proteins was also checked with serum recognizing Om44, a P-selectin salivary protein from O. moubata whose neutralization induces antibody block tick feeding. To test the functionality of the proteins in the array, they performed protein-protein interaction studies with the recombinant P-selectin/Fc chimera. With this aim, the proteins on the array and P-selectin/Fc chimera were expressed in situ normally and also in the presence of canine pancreatic microsome membranes (CMMs). They found that P-selectin/Fc chimera interacted with phospholipase A2 (PLA2) expressed in situ on the array. This finding suggested that this secreted O. moubata PLA2 (sPLA2) could be a potential P-selectin interacting partner [11].
As another example, a NAPPA array was designed for systematic characterization of viral protein-host interactions. Through the access to viral ORFs in flexible cloning formats, the LaBaer’s lab is releasing the initiation of a panviral proteome collection of 2035 ORF clones from 830 viral genes in the Gateway® recombinational cloning system. In this work, NAPPA arrays are suitable, highly efficient and flexible platforms for displaying viral proteins and detecting host serological responses using micro-fluidic multiplexed immunoassays and allowing the study of host-viral protein interactions [12]. Related to host-pathogen interactions in Legionella pneumophila infections, this group have applied NAPPA technology to determine the interaction network of the pathogen with 10,000 unique human proteins. They identified known and novel interaction candidates and, additionally, substrates for an effector with and adenylyl transferase domain that catalyzes AMPylation. Their results highlighted the amenability of NAPPA to high-throughput analysis of effectors from a wide variety of human pathogens [13].
Nicolini and collaborators clinically screened neuro-oncological patients respondent to temozolomide (TMZ) from those showing resistance to the drug by using a NAPPA-based nanoconductometric sensor [14]. Their results shower a properly discrimination of protein-protein interactions depending on the behavior against TMZ [15]. Finally, Liang et al. have successfully coupled two different technologies (label-free and real-time detection method plasmonic-based electrochemical impedance microscopy with NAPPA arrays) to determine small molecule binding kinetics. This approach allowed the measurement of binding kinetics and affinity parameters between small molecule drugs (imatinib and SB201290) and their target proteins (kinases ABl1 and p38-α) with high sensitivity and reproducibility. These results demonstrate that NAPPA methodology is a reliable technology to understand small molecules interactions in biological systems and is also useful in the discovery of small molecules drugs [4].

3.2. Vaccine Development by NAPPA Technology
Since the development of NAPPA arrays, many research groups have used different NAPPA-based platforms for investigating immune diseases and improving vaccine development. Next, we briefly describe some of these studies.
Respiratory tract and lung infections in cystic fibrosis (CF) patients and individuals who are otherwise immune compromised can be caused by a gram-negative bacterium called Pseudomonas aeruginosa. This microorganism was selected by Montor et al. [16] to test candidate membrane antigens with NAPPA arrays. The principal goal of their work was to map the immune responses of patients infected with P. aeruginosa to determine which bacterial outer membrane proteins induced a strong immune response. The principal difficulty in purifying membrane proteins to display on NAPPA arrays is related to their hydrophobic domains. They designed a NAPPA array containing all 262 outer membrane proteins of the bacterium. Serum samples from 22 CF patients with documented pseudomonal pneumonia and 16 non-CF individuals with various acute P. aeruginosa infections as well as 15 healthy controls were selected for array screening. After analysis, 12 proteins triggering an adaptive immune response were identified in a majority of the infected patients, yielding valuable information about which bacterial proteins are recognized by the immune system during the natural course of infection.
In turn, Ceroni and colleagues [17] used NAPPA arrays for a systematic analysis of the IgG antibody immune response against varicella zoster virus (VZV), a human herpes virus, encoding at least 69 distinct viral proteins, which causes chickenpox after primary infection and shingles during reactivation. Its effects are particularly important in pregnancy and immunocompromised patients and sera-diagnostic tests are commonly used for its detection. In order to investigate the humoral immune response to VZV infection or vaccination in more detail, Ceroni developed a specific NAPPA array containing all 69 VZV proteins mentioned above and performed a detailed analysis of 68 sera from individuals with either no infection or an acute VZV infection. The obtained results confirmed previous knowledge about viral open reading frames (ORF) such as reactive glycoproteins antigens (ORF 5, ORF 14, ORF 31, ORF 37, ORF 68), and also found novel responses against a variety of other membrane proteins (ORF2, ORF24), capsid (ORF20, ORF23, ORF43) and tegument (ORF53, ORF9, ORF11), as well as others related to virus replication (i.e., ORF 25, ORF26, ORF28) and transactivator proteins (ORF12, ORF62 and ORF63).

3.3. Evaluation of Autoimmune Responses
In autoimmune diseases, antibodies, known as “auto-antibodies,” are often generated by the humoral immune system against self-proteins in response to many pathological processes. This kind of antibody follows a specific pathway to recognition by the immune system, including antigen over-expression, mutation, and/or altered PTM released from damaged tissues [18,19]. The presence of these autoantibodies is related to the development of certain diseases such as diabetes. Thus, they can be useful as diagnostic/prognostic biomarkers [20]. As diagnostics they have several key advantages: (i) they can be detected even before the appearance of clinical symptoms; (ii) even if the antigen that induced them is absent or present at very low amounts, the antibodies can be readily detected; (iii) they can be measured from easily obtained sources such as blood other body fluids; (iv) they are very stable in standardized collection vessels; and (v) they are straightforward to measure using many available chemistries.
In 2007 Anderson and colleagues employed NAPPA arrays for serological screening in breast cancer. After NAPPA design, they tested p53 together with other three negative control antigens (S100A7, p21 and ML-IAP) with positive and negative p53 sera confirming the expression for all the proteins printed and checking the detection of antibodies against p53. Moreover, they determined differences in p53-expression levels between healthy donors and breast cancer patients, and also within disease stages. In addition, they confirmed that many regions of the protein expressed were accessible on the arrays. To extend the study to autoantibody biomarker detection, they built a high density NAPPA array printing 1117 cancer related genes of which 539 were implicated on breast cancer and tested them against melanoma, ovarian and breast cancer sera [21]. Later, they increased the number of novel autoantibodies to be tested in breast cancer (around 4988 candidate antigens). Finally they identify 28 autoantibodies that could distinguish between benign breast disease and invasive cancer in a blinded study [22].
Recently, LaBaer et al. developed several NAPPA studies for juvenile idiopathic arthritis (JIA) [23] and type 1 diabetes (T1D). In both cases, serological autoantibodies (AAbs) from the disease were screened using a two-stage method. Firstly, more than 6000 unique proteins were displayed in NAPPA arrays which were incubated with 50 sera from T1D patients and 20 from controls allowed the elimination of uninformative antigens. In the second stage, 750 remaining genes were printed in duplicate and 26 proteins were identified as novel AAbs (TBCA, CDK4, CDK6, TBRG4, among others) with p < 0.005 [24]. For juvenile idiopathic arthritis, they assessed the levels of antibodies present in the systemic circulation and synovial joint of a small cohort of juvenile arthritis patients as a pilot study. Their results showed a strong correlation between the circulating antibody levels and those of the inflamed joint.
In 2009, Wong et al. adapted NAPPA technology to the Luminex suspension bead array platform to monitor the humoral immunity. To accomplish this, they expressed the proteins and captured them with the Luminex beads through anti-tag antibodies. After mixing the antigen-loaded beads, serum was added and human IgG was detected with standard secondary detection reagents. Protein arrays are a useful method for testing a moderate number of clinical samples against thousands of candidate proteins. The advantage of the Luminex approach is that it allows testing hundreds of clinical samples against a moderate number of candidate antigens, i.e., mesoscale. They concluded that detection of antibodies was highly reproducible and the specificity and limits of detection of the platform were comparable to standard ELISAs [25].
Recently, Henjes and Lourido have performed an analysis of auto-antibody profiles in osteoarthritis (OA) using comprehensive protein arrays concepts. In this work, NAPPA arrays and antigen arrays have been used to characterize differential autoantibody profiles in a set of 62 samples from OA, rheumatoid arthritis (RA), and healthy controls. An untargeted screen was performed on 3840 protein fragments spotted on planar antigen arrays, and 373 antigens were selected for validation on bead-based arrays. In the NAPPA approach, a targeted screening was performed on 80 preselected proteins. The autoantibody targeting CHST14 was validated by conventional ELISA assays in the same set of patients. Altogether, nine and seven disease-related autoantibody target candidates were identified, respectively, and this work demonstrated a combination of these two array concepts for biomarker discovery and their usefulness for characterizing disease-specific autoantibody profiles [26].