Development of high-capacity affinity membranes for protein isolation requires membrane pores coated with thin films that bind multilayers of proteins. To prepare membranes that selectively capture polyhistidine-tagged (His-tagged) proteins, this work explores layer-by-layer adsorption of polyelectrolytes containing chelating groups that form Ni2+ complexes. Sequential adsorption of protonated poly(allylamine) (PAH) and carboxymethylated branched polyethyleneimine (CMPEI) leads to membranes that bind Ni2+ and capture ~60 mg of His-tagged ubiquitin per mL of membrane. Both binding capacity and metal-ion leaching are similar to values seen with high-binding commercial beads, but membranes should facilitate protein isolation in minutes.   Membranes are also convenient substrates for trypsin immobilization and subsequent proteolysis. Passage of protein solutions through 100-m thick trypsin-modified membranes enables reaction residence times as short as milliseconds to limit digestion and provide large peptides for mass spectrometry (MS) analysis. Large peptides can both enhance protein sequence coverage and help identify flexible regions in a protein. With either cytochrome c or apomyoglobin, in-membrane trypsinolysis cleaves the protein after lysine residues in highly flexible regions to generate two large peptides that cover the entire protein sequence.       iv   v         vi     ...............................................................................................    ..........................................................................................   ......................................................................................    ...........................................................................    ...........................................................................  ...........................................................................................................   ................................................................................................   ........................................   ...............................................   ..........................................   .........................................................................................................   .....................................................................................   .......................................   .......................   ..........................................................................   ......................................................................................................................   .....................................................   ..............................................................................   ......................................   .................................................   ...............................................................................   .............................   .............................................................................................   .....................................................................................................................    ................................................................................................................   ....................................................................................................................   ......................................................................................................   ..................................................................................................................   ......................................   .................................................   .......................................   ......................................................................................   ....................................................................  vii   ..............................................   ....................................................................................................   ....................   ..........................................................   ...........................................................................................................   .........   .......................................................................................................   ....................................................................................................   .................................................................................................................................   ..................................................................................................   .................................................................................   ....................................................................................................................................   .....................................................................................................................................   ........  ....................................................................................................................   ..................................................................................................................   ..................................................................................................................   ..............................................................................................................................   .............................   .............   ..................................................................................................   .................................................................................................................   ............................   .....................................................................................................   ...................................................   .........................................................................     ..........................   ...................................................................................................   .................   ............................................................   ...................................................   .........................................................................................................   ....   ........................................   ....................................................................................................................   .....................................................................................................................  viii   ...........................................................................   ..................................................................................................................   ....................................................................................................   ................................................................................................................   ..................................................................   ..........................................................   .................................................................   ..............................................   .....................................................   .....................................................................   ..................................................................................................   .....................   .............................................................   ...................................................................................................   ..   .................................   ....................................................................................................................   ...................................................................................................................    ......................................................................   ..................................................................................................................   ....................................................................................................   ................................................................................................................   .......................................................................................   ..........................................................................................................   .....................................................................   ....................................................   ..................................................................................................   ..............................................................................................   .................................................................   ...................................................................................................   ...................................................................................................................................   ........................................................................................   .........................................................................   ..................................................................................................................   ...................................................................................................................    .........................................................   .......................................................................................................   .   ...............................................  ix   .................................   ............   ........................................................   ............................................................................................................   ...................................................................................................................    x  LIST OF TABLES  .....................................................................................................................................    ........................................................................    .............   ...........................................................................................    .........................................................................................................................................    ...........................................................................................    .................................................................................................................    xi  LIST OF FIGURES  ............................................................................................................  ....................................................................................................................................................   .........................................   ............................................................................................................................   .......................................   .......................................................   ............................................................................................................  .......................................................................................................................................   ...........................................................................................   ...................................................................................   ...................................   .........................   ...........................................................  xii   ...........................................................................   ............   .................................................................................   .............................................   ......................................................   ................................................................................................................................   ................................................................................................   ..................................................   ...............................................................................  xiii  .................  ..................................................................................................................................................   ..................................................   .....................................................................................................   ..............   ....................   ............................................................................................................................   ....................................  xiv  .....................................................   .................................................................................................................................  ............................................................   .....   ....................................................................................................................................   ......................................................................................................................   ........................   ....................................  xv   ....................................................................................................................................   ......................................................................................   ...........................................................................................................   ............................................................   .................................................................................................................................   ..............................................................................................................................   ......................................................................................  xvi   ......................................................................................................................................   ..............................................................................................................................   .................................   ...................................................................   ...................................................................................................................   .....................................................   ...........................................................................   ..............................................................................................................................     .............................................................................................   .................................................................................  xvii   ..............................................................................................................................  ................................................................................................................................................   ...........................................................................................................................   ......................................................................................................   ...................................................................................   .......................................................................................   .............................................................................................................  xviii   ................................................................................................................   ....................................................   ....................................................   .......................................  ................................................................................................................................................  ................................................................................................................................................   ............................................    xix  LIST OF SCHEMES  ..................................................................   .........................................................................................................   ................................................................................  ................................................................................................................   ..    xx  KEY TO ABBREVIATIONS                                                 xxi                                                  xxii                     1 Chapter 1. Introduction     2       3     4       5                                           6      7  Both surface-initiated growth of polymer brushes and layer-by-layer (LbL) polyelectrolyte adsorption can provide highly swollen films that capture multiple layers of proteins, .30-37 Compared to the synthesis of polymer brushes, which is a relatively cumbersome process that frequently requires initiator immobilization and subsequent polymerization under anaerobic conditions, LbL deposition is quite simple. Our group employed LbL adsorption of poly(acrylic acid) (PAA)/(polyethyleneimine) (PEI) films followed by derivatization with aminobutyl nitrilotriacetate (NTA) and Ni2+ to form NTA-Ni2+ complexes that capture His-tagged proteins.38   Moreover, in addition to NTA these membranes contain residual -COOH groups of PAA that bind metal ions only weakly, which likely increases metal-ion leaching.  Thus, examining whether direct adsorption of relatively inexpensive polyelectrolytes containing chelating groups can effectively modify membranes to  8 bind metal ions and capture His-tagged protein is potentially important to further decrease the cost for His-tag protein purification. Chapter 2 describes our efforts in this area.        9       10                          11                                     12     13     14     15     16      17     18                          19      20       21                        22     23       24      25                      26                                           27                                       28                                   29                               30                                     31                                     32                                  33                        34 Chapter 2. Immobilization of Carboxymethylated Polyethyleneimine-Metal Ion Complexes in Porous Membranes to Selectively Capture His-tagged Protein    In most studies of overexpressed proteins, purification employs engineered affinity tags.1 Hexahistidine is the most common affinity tag because it is relatively small and enables convenient capture by binding to beads containing Ni2+ or Co2+ complexes.2,3 Nevertheless, bead-based separations suffer from slow diffusion of large macromolecules into nanopores,4-7 which necessitates long separation times that may harm sensitive proteins.  Purifications are especially time consuming when capturing proteins from large volumes of dilute solutions.  Porous membranes modified with affinity ligands are an attractive alternative purification platform because convection through the membrane pores and short radial diffusion distances provide rapid protein transport to binding sites.8,9 Moreover, membrane pressure drops are low because of small thicknesses.10-14 However, membranes have a lower specific surface area than nanoporous beads, which often leads to a low binding capacity.   To increase protein-binding capacities, several groups modified membrane pores with thin polymer films.  Both surface-initiated growth of polymer brushes and layer-by-layer (LbL) polyelectrolyte adsorption can provide highly swollen films that capture multiple layers of proteins.5,15-21 Compared to the synthesis of polymer brushes, which is a relatively cumbersome  35 process that frequently requires initiator immobilization and subsequent polymerization under anaerobic conditions, LbL deposition is quite simple. Our group employed LbL adsorption of poly(acrylic acid) (PAA)/(polyethyleneimine) (PEI) films followed by derivatization with aminobutyl nitrilotriacetate (NTA) and Ni2+ to form NTA-Ni2+ complexes that capture His-tagged proteins.22   Moreover, in addition to NTA these membranes contain residual -COOH groups of PAA that bind metal ions only weakly, which leads to metal-ion leaching.   This study examines whether direct adsorption of relatively inexpensive polyelectrolytes containing chelating groups effectively modifies membranes to bind metal ions and capture His-tagged protein (Figure 2.1).  Specifically, we adsorb protonated poly(allylamine) (PAH)/ (PDCMAA) or PAH/carboxymethylated branched polyethyleneimine (CMPEI) films in membrane pores in ~40 minutes.  Both PDCMAA and CMPEI contain iminodiacetic acid groups that form during reaction of the commercial polymers PAH or branched PEI with sodium chloroacetate (Scheme 2.1).23 Thus, these polymers are readily accessible synthetically and relatively inexpensive.  Previous studies examined LbL adsorption of (PAH/PDCMAA)n films and showed that they can contain up to 2.5 M of metal ions and facilitate selective metal-ion transport.24,25  Carboxymethylated linear PEI is commercially available, but we employ branched PEI because it may provide thicker, highly swollen films for protein capture.26 Importantly, we compare protein binding to PAH/PDCMAA and PAH/CMPEI films to test our hypothesis that ammonium groups in the PEI backbone will increase swelling and enhance  36 protein capture. Membranes modified with PAH/CMPEI rapidly capture as much as 60 mg of protein per mL of membrane, which is equivalent to the capacities of high-binding commercial beads.27,28     Aqueous solutions containing 0.02 M PAH, 0.01 M CMPEI or 0.01 M PDCMAA were prepared in deionized -Q) or 0.5 M aqueous NaCl, and solution pH values were adjusted by dropwise addition of 0.1 M NaOH or HCl. Polymer concentrations are given with respect to the repeating unit.  Au-coated Si wafers (200 nm of sputtered Au on 20 nm of Cr on Si (100) wafers) were cleaned in a UV/O3 chamber for 15 min prior to use. Other m thick), Conconavalin A (Con A from Canavalia ensiformis (Jack bean), Sigma Aldrich), coomassie protein assay reagent (Thermo Scientific), histidine6-tagged ubiquitin (His-U, human recombinant, Boston Biochem), poly(allylamine hydrochloride) (Mw Alfa Aesar), and poly(sodium 4-styrenesufonate) (PSS, Mw ~ 70 000 Da, Sigma Aldrich). CMPEI synthesis employed a branched poly(ethyleneimine) solution (Mn ~60,000 Da by gel-permeation chromatography, average Mw ~750,000 Da by light scattering, 50 wt. % in H2O,      37 Sigma Aldrich). Cupric sulfate, nickel sulfate, sodium phosphate, sodium phosphate dibasic, ethylenediaminetetraacetic acid disodium salt (EDTA), sodium chloroacetate (98%), 3-mercaptopropionic acid (MPA, 99%) and imidazole (>99%) were received from Aldrich and used without further purification. Buffers include:  binding buffer 1:  20 mM phosphate, pH 6; binding buffer 2:  20 mM phosphate, pH 7.4; washing buffer 1:  20 mM phosphate, 150 mM NaCl, 0.1% Tween 20, pH 7.4; washing buffer 2:  20 mM phosphate, 45 mM imidazole, 150 mM NaCl, pH 7.4; elution buffer:  20 mM phosphate, 500 mM NaCl, 500 mM imidazole, pH 7.4; stripping buffer:  20 mM phosphate, 500 mM NaCl, 50 mM EDTA, pH 7.4. Unless noted otherwise, uncertainties are standard deviations of values derived from three experiments with independent membranes or wafers.  The synthesis of PDCMAA was published previously,23,25 and synthesis of CMPEI was carried out following the procedure for synthesis of PDCMAA, with slight modifications (Scheme 2.1).  (Syntheses were performed by Salinda Wijeratne.)  Under a N2 atmosphere, sodium chloroacetate (20.0 g, 0.25 mol) and 25 mL of water were added to a two-neck round-bottomed flask, and the mixture was stirred at 30 °C for 10 min. This solution was added dropwise with stirring to an aqueous solution (100 mL) containing poly(ethyleneimine) (PEI, 50 wt% solution in water, Mn~6.0 x104 Da, 10.0 g, 10.6 mmol, assuming a repeating unit MW=473 gmol-1) at 50 °C. The reaction mixture was kept at 50 °C for 1 h and then held at 90 °C for 2 h with occasional addition of 30% NaOH to maintain the pH at 10.0. The mixture was stirred at room temperature for 12 h, and then the pH was adjusted to 1 by adding concentrated HCl. The supernatant was decanted, the remaining precipitate was dissolved by addition of  38 30% NaOH, and the solution was again adjusted to pH 1.0 with concentrated HCl. This process was repeated 3 times, and the precipitate was filtered and dried in vacuo for 12 h. The resulting white carboxymethylated polyethyleneimine (CMPEI, solid, 3.2 g, 63% yield) was characterized by FTIR spectroscopy (KBr) and elemental analysis.              To assess the acid-base properties of the new polymer, 30 mL of 1 mg/mL CMPEI (dissolved in 0.025 M NaOH) was titrated with 0.1 M HCl using a pH meter (ORION-420A). The pH meter has a combined glass/reference electrode and was calibrated with pH 4, pH 7, and pH 10 standards. A volumetric pipette was used to add HCl to the CMPEI solution with vigorous stirring.  39  Au-coated Si substrates (24 mm × 11 mm) were immersed in 5 mM MPA in ethanol for 16 h, rinsed with ethanol, and dried with N2 to form a monolayer of MPA prior to adsorption of PAH. The MPA-coated substrates were immersed in 0.02 M PAH (adjusted to the desired pH) for 15 min and subsequently rinsed with 10 mL of deionized water and blown dry with N2. Substrates were then immersed in a 0.01 M CMPEI or PDCMAA solution (adjusted to the desired pH value) for 15 min followed by the same rinsing and drying procedures. Adsorption presumably displaces counterions from the polyelectrolytes and creates electrostatic cross-links between PAH and CMPEI or PDCMAA to stabilize the films despite the high water-solubility of these polymers.29  In some cases, the polyelectrolyte solutions also contained 0.5 M NaCl. The process was repeated to form multilayer films.  For some experiments, nylon membranes were first immersed in 0.1 M sodium chloroacetate in 3 M NaOH for 16 h and subsequently washed with deionized water and dried with N2. The resulting carboxymethylated membrane disks were cleaned for 10 min with UV/O3 and placed in a homemade Teflon holder (similar to an Amicon cell) that exposed 3.1 cm2 of external membrane surface area. The UV/O3 exposure should oxidize contaminants or the surface of the membrane but have minimal effect on the membrane structure.30  Subsequently, a 5-mL solution containing 0.02 M PAH and 0.5 M NaCl was circulated through the membrane for 15 min at a flow rate of 1 mL/min using a peristaltic pump. A CMPEI or PDCMAA layer was deposited similarly using 0.01 M CMPEI or 0.01 M PDCMAA solutions containing 0.5 M NaCl. After deposition of each polyelectrolyte layer, 20 mL of deionized  40 water was passed through the membrane at the same flow rate. Nylon membranes without carboxymethylation were modified with PEMs similarly, starting with the UV/O3 cleaning.    Spectroscopic ellipsometry (model M-44; J.A. Woollam) was used to determine the thicknesses of PEMs on gold-coated Si wafers, assuming a film refractive index of 1.5. Film thicknesses in aqueous solutions were measured in a home-built cell described previously.31 In that case, the software determines the refractive index of swollen films. Reflectance FTIR spectra were obtained with a Thermo Nicolet 6700 FTIR spectrometer using a Pike grazing angle (80°) apparatus. A UV/O3-cleaned Au-coated Si wafer served as a background.   Bare carboxymethylated membranes and membranes modified with (PAH/CMPEI)n and (PAH/PDCMAA)n films were loaded with Cu2+ or Ni2+ by circulating 5 mL of 0.1 M CuSO4 or NiSO4 (pH  4 for both) through the membrane for 30 min, followed by passage of 20 mL of water through the membrane. Metal ions were eluted from the membranes with 5 mL of stripping buffer or 2% HNO3 and subsequently analyzed by atomic absorption spectroscopy. For protein capture on wafers coated with PEMs, the modified substrates were immersed for 1 h in solutions containing 0.3 mg/mL of Con A in binding buffer 1 or 0.3 mg/mL of His-U in binding buffer 2. Subsequently, using a Pasteur pipet these substrates were rinsed with 10 mL of washing buffer 1 and 10 mL of water for 1 min each and dried with N2. The amount of protein binding was determined by reflectance FTIR spectroscopy and expressed as the equivalent thickness of spin-coated protein that would give the same absorbance. The  41 1 32 Some of these thicknesses were confirmed using ellipsometry. If the protein density is 1 g/cm3, each nm of equivalent thickness is equal to 1 mg/m2 of surface coverage.  Protein breakthrough curves were obtained by passing protein solutions (0.3 mg/mL in binding buffer 1 or binding buffer 2) through the membranes. For Con A binding, these studies employed 3.1 cm2 of external membrane surface area.  His-U binding experiments used a Teflon holder that exposed a membrane area of 0.78 cm2 (1.0-cm exposed diameter) because of the high cost of this protein. Bradford assays (using calibration with the protein of interest) were employed to quantify the concentrations of proteins in the membrane effluent or eluate.   His-tagged small ubiquitin modifier (His-SUMO) was overexpressed in Escherichia coli (E. coli) cells. The cells were lysed with sonication in binding buffer 2 and centrifuged. Supernatant was pumped thorough the (PAH/CMPEI)-modified membrane (diameter 2-cm) at room temperature at a flow rate of 1 mL/min. Subsequently the membrane was rinsed with 5 mL of binding buffer 2 and 5 mL of washing buffer 1, and the bound protein was eluted with 2 mL of elution buffer. The purity of the eluted protein was determined by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDSPAGE).   To test metal-ion leaching in different buffers, (PAH/PDCMAA)-, (PAH/PDCMAA)2-, (PAH/CMPEI)- and (PAH/CMPEI)2-modified carboxymethylated nylon membranes were loaded with Ni2+ using the above procedure (including rinsing with 20 mL of water) and washed  42 consecutively with 160 bed volumes (5 mL) of binding buffer 2, washing buffer 1, washing buffer 2, elution buffer, stripping buffer, and 2% HNO3. As a comparison, a GE Healthcare HiTrapTM IMAC FF column (1 mL) was washed with 160 bed volumes (160 mL) of the same buffers. All the samples were diluted 1:5 with deionized water and analyzed by atomic absorption spectrometry. The GE Healthcare HiTrapTM IMAC FF column was loaded with Ni2+ by passing 2 mL of 0.1 M NiSO4 through the syringe column (flow rate of 1 mL/min) followed by 160 mL of deionized water.   To examine film stability under purification conditions, we soaked (PAH/CMPEI)2-modified gold wafers in 5 mL of binding buffer 2 for 20 hours. Film thickness values and reflectance FTIR spectra were obtained before and after immersion in the buffer for different times. Total organic carbon (TOC, O.I. Analytical, Model 1010) analysis was used to quantify polyelectrolyte leaching from modified membranes during passage of binding buffer 2 through the membrane. CMPEI solutions with concentrations from 0 to 10 ppm were used for calibration, and the effluent was diluted 1:39 with deionized water before analysis. To study reusability, multiple cycles of charging with Cu2+, binding of Con A, rinsing, and elution were performed with a (PAH/CMPEI)-modified membrane (deposited at pH 2 with 0.5 M NaCl).  Protein binding was calculated from the average of capacities determined from the breakthrough curve and the eluate analysis.        Comparison of the IR spectra (Figure 2.2) of acidified branched PEI and branched CMPEI shows the disappearance of bands that correspond to N-H deformation vibrations of PEI (1584 cm-1 and 1454 cm-1) and the appearance of stretches from carboxyl groups. The absorption at  43 1733 cm-1 arises from the C=O stretching in the HN+-CH2COOH group, and the band at 1655 cm-1 is due to the asymmetric stretching in the HN+-CH2COO- group. This confirms the presence of the iminodiacetic moiety in CMPEI.   . KBr FTIR spectra of branched PEI (red) and CMPEI (black).  Both polymers were acidified prior to obtaining the spectra.  We also performed elemental analysis to evaluate the synthesis of CMPEI.  Table 2.1 provides possible structures for PEI and CMPEI along with elemental compositions. In its fully deprotonated state, the PEI starting material has the following percent composition:  C, 55.78; H, 11.70; N, 32.52. Double carboxymethylation of each primary amine plus single carboxymethylation of each secondary amine leads to entry 2 (Table 2.1) with a percent composition of C, 47.52; H, 6.98; N, 13.85; O, 31.65.   However, these values differ significantly from the CMPEI experimental elemental analysis data:  C, 40.26; H, 6.65; N, 11.93.  This difference likely stems from formation of hydrochloride salts. Without          44 accounting for chloride, all other atom percentages will be artificially high.  Formation of HCl salts only along the polymer backbone (addition of 5 Cl-) gives an elemental composition:  C, 40.83; H, 6.39; Cl, 13.69; N, 11.90; O, 27.19 (Table 2.1, entry 3).  This is reasonably close to the experimental values.  Formation of HCl salts at all amine sites (Table 2.1, entry 4) leads to atomic percentages that are significantly lower than the experimental values. Unfortunately, we cannot specify the protonation state of CMPEI because of the low COOH pKa values, and COO- groups, rather than Cl-, probably provide charge compensation for some of the ammonium groups.  Thus, 5 Cl- ions per repeating unit, as shown in entry 3 of Table 2.1, is possible.  Most important, in entries 2-4 the carbon to nitrogen ratio, which does not depend on the number of Cl- ions or the presence of residual water, is 3.43 close to the experimental value of 3.37.  This confirms addition of acetate groups to the polymer in approximately the amount shown in entry 2.  (The theoretical C to N ratio in the PEI starting material is only 1.72.)    45 .  Possible elemental compositions of PEI and CMPEI with different numbers of HCl salts.   Chemical  Formula Molecular Weight Elemental Analysis 1  C22H55N11 473.8 C, 55.78; H, 11.70; N, 32.52 2  C44H77N11O22 1112.1 C, 47.52; H, 6.98; N, 13.85; O, 31.65 3  C44H82Cl5N11O22 1294.5 C, 40.83; H, 6.39; Cl, 13.69; N, 11.90; O, 27.19 4  C44H88Cl11N11O22 1513.2 C, 34.93; H, 5.86; Cl, 25.77; N, 10.18; O, 23.26    46  CMPEI contains both weakly basic (amine) and weakly acidic (carboxylic acid) groups and thus can potentially form salt bridges with both cations and anions on a surface. Figure 2.3a shows an acid titration of CMPEI. The number of protonated equivalents in CMPEI (Figure 2.3b) was calculated using the following equations:  H+(added from titrant) =H+(free in solution) + H+(added to CMPEI) + OH-(neutralized)                                                    (1)        (2) In equation (2), 1112 is the molecular mass of the CMPEI repeat unit in Table 2.1, entry 2, and there are 11 amine groups in the CMPEI repeat unit.  This calculation may underestimate the number of equivalent by ~15% if there are 5 chloride ions in the solid polymer. However, the trend in the figure should hold.   Figure 2.3 suggests nearly complete protonation of amine groups at pH values below 7, whereas protonation of the carboxylate groups begins below pH 4, which is similar to the titration of PDCMAA.25 This is reasonably consistent with the pKa values for iminodiacetic acid, which are 9.4, 2.6, and 1.8.33,34 The ratio of carboxylic acid groups to amines is around 1:1 in CMPEI but 2:1 in iminodiacetic acid and PDCMAA (see Scheme 2.1).   47   . (a) Titration curve for 1 mg/mL of CMPEI in 0.025 M NaOH (30 mL). The titrant contained 0.1 M HCl; (b) Number of Equivalents (the number of protons added to CMPEI divided by the number of CMPEI amine groups) as a function of pH.  Based on the polymer titration and a 1:1 ratio of amine to carboxylic acid groups, one might suppose that CMPEI would serve as a polyanion in films formed at basic pH and as a polycation in films formed at acidic pH.  However, Hoffman and Tieke reported that linear CMPEI, which also has a 1:1 ratio of amine to carboxylic acid groups, forms multilayer films with protonated poly(vinyl amine) at adsorption pH values ranging from 2 to 8.35  Thus, even at pH 2, linear CMPEI likely serves as a polyanion in LbL deposition.  With branched CMPEI,  48 adsorption of (polycation/CMPEI)n coatings also occurs at low pH. Figure 2.4 shows the ellipsometric thicknesses of (PAH/CMPEI)n films deposited at pH 3.  In the absence of salt in adsorption solutions (red circles), after deposition of the initial bilayer, which is ~1 nm thick, adsorption of each subsequent bilayer adds ~5 nm of thickness. Addition of 0.5 M NaCl to adsorption solutions increases the thicknesses of layers 2- to 4-fold.  At low pH, CMPEI has a net positive charge, so electrostatic repulsion between its positive ammonium groups should make the polymer chains partially extend. Addition of salt increases thickness by screening charges in the polymer to create loops and tails and by increasing surface roughness.36,37    . Ellipsometric thicknesses of (PAH/CMPEI)n films as a function of the number of adsorbed bilayers, n. Films were deposited from pH 3 solutions containing 0.5 M NaCl (blue squares) or no added salt (red circles). The substrates were Au-coated Si wafers modified with a monolayer of MPA, and error bars are typically smaller than the symbols. During adsorption, carboxylate groups on CMPEI most likely bind to ammonium groups of PAH. Reflectance FTIR spectroscopy confirms that most of the carboxylate groups in these films are deprotonated (Figure 2.5). Formation of PAH/CMPEI complexes leads to less protonation of the COOH groups of CMPEI than in solution and perhaps less protonation of ammonium groups. XPS data (Figure 2.6) show no chloride within CMPEI-capped films, 051015202530354045500246Thickness (nm)Number of bilayers (n) 49 which suggests that few of the amine groups in CMPEI are protonated and compensated by Cl- ions. The formation of films by adsorption of CMPEI and PAH, which both possess a net positive charge in neutral and acidic solutions, likely occurs due to polarization-induced attraction.38-40 Electric fields created by positively charged PAH may induce rearrangement of the CMPEI chains to enhance electrostatic interactions between the carboxylates of CMPEI and ammonium groups of PAH. At pH 3 with 0.5 M NaCl, (PAH/CMPEI) growth reaches a plateau at 4-5 bilayers, perhaps because the net positive charge on both polymers leads to repulsions that overcome polarization-induced attraction in thicker films.   . Reflectance FTIR spectra (2200-800 cm-1) of (PAH/CMPEI)5 films deposited at pH 2, 3, 5, 7 or 9 on MPA-modified, Au-coated Si wafers.            50  . XPS spectrum of a (PAH/CMPEI)5 film adsorbed on a MPA-modified, Au-coated Si wafer.  Deposition occurred at pH 2 with 0.5 M NaCl in the polyelectrolyte solution, and the wafer was rinsed extensively with water. Figure 2.7 shows the thicknesses of (PAH/CMPEI)5 films as a function of the deposition pH. Similar to other films with weak-acid polyelectrolytes,25,41,42 the highest thicknesses occur with films deposited at the lowest pH.  Films formed at pH 2 are typically about twice as thick as films adsorbed at pH 3-9.  Due to the relatively low pKa values of the COOH groups in CMPEI, thickness only increases at the lowest pH value.  Notably, 4- and 5-bilayer films deposited at pH 3 are thinner than corresponding films deposited at all other pH values (compare Figures 2.4 and 2.7).  This may reflect repulsion between CMPEI and PAH at this pH.  At pH 2, an increased number of protonated COOH groups may require more CMPEI                    51 to form ion pairs with PAH and overcome decreases in thickness due to repulsion between the two polymers.    . Ellipsometric thicknesses of (PAH/CMPEI)5 films as a function of deposition pH.  Films were adsorbed from 0.5 M NaCl solutions onto Au-coated Si wafers modified with a monolayer of MPA, and error bars are often smaller than the symbols.  (For coatings adsorbed at pH 2, non-integer bilayer numbers indicate films terminated by PAH adsorption.) For the pH 2 deposition, we also determined the thickness increases due to adsorption of both PAH and CMPEI. As Figure 2.7 shows (blue squares), the thickness increase upon adsorption of CMPEI is more than double that for adsorption of PAH, suggesting that the films contain more CMPEI than PAH, probably because the density of COO- groups on CMPEI is lower than the density of protonated amine groups on PAH. After deposition of the fifth (PAH/CMPEI) bilayer at pH 2, the surface is too rough for an accurate thickness determination by ellipsometry.  The reflectance IR spectra of (PAH/CMPEI)5 films deposited at different pH values show that most of the carboxylic groups are deprotonated (Figure 2.5). However, the ratio of the absorbance of the -COO- stretch (~1650 cm-1) to the absorbance of the acid carbonyl stretch 04080120160200012345Thickness (nm)Number of bilayers (n)pH 2pH 5pH 7pH 9 52 (1720 cm-1) decreases as the deposition pH decreases, suggesting the films deposited at the lowest pH values contain free -COOH groups. Figure 2.9 shows the reflectance FTIR spectra of films with 1 to 5 (PAH/CMPEI) bilayers for different deposition pH values.     53    54 CMPEI gives very thin films when serving as a polycation in LbL adsorption. (CMPEI/PSS)5 films deposited at pH 3 in 0.5 M salt are only 10±2 nm thick. The positive charges of CMPEI reside mostly in or near the backbone and may be less available for adsorption than -COO- groups on the side chains.  Using a cyclic analogue of linear CMPEI, Hoffman and Tieke also found minimal growth during LbL deposition with PSS over a pH range from 2 to 8.35  This work aims to create thin films that selectively bind proteins in platforms such as porous membranes, and film swelling in aqueous solution is vital to enable extensive protein capture. To examine swelling, we initially performed in situ ellipsometry with (PAH/CMPEI)5 films (deposited at pH 3 with 0.5 M NaCl) immersed in deionized water or binding buffer 2 (pH 7.4). After a 20-minute immersion, film thickness increased 160±30% in deionized water and 680±260% in buffer. Consistent with the approximately 62% and 88% water in the immersed coatings, the film refractive indices decrease from 1.50 to 1.39 and from 1.50 to 1.35 after swelling in water and buffer, respectively. (The refractive index of water at the wavelengths of the spectroscopic ellipsometer is about 1.333.) Deprotonation of carboxylate groups in pH 7.4 buffer likely enhances swelling, which should provide space for binding multilayers of protein in the film.  IR spectra confirm the deprotonation after immersing the film in buffer (see Figure 2.9). As a comparison, the swelling of (PAH/PDCMAA)5 films (deposited at pH 3 with 0.5 M NaCl) was 52±16% in deionized water and 220±20% in binding buffer 2. The high swelling of (PAH/CMPEI)5 relative to (PAH/PDCMAA)5 suggests that the ammonium-containing backbone and branched structure of CMPEI facilitate swelling.   55 ((PAH/CMPEI)5 and (PAH/PDCMAA)5 films have similar dry thicknesses of 40 and 60 nm, respectively.)  Note that high swelling may lead to partial polyelectrolyte desorption, which we discuss in section 3.8.   . Reflectance FTIR spectra (2200-800 cm-1) of a dry (PAH/CMPEI)2 film after adsorption and rinsing with water (black) and after immersion in binding buffer 2 (pH 7.4) followed by rinsing with water (red). The film was initially adsorbed on a MPA-modified, Au-coated Si wafer at pH 3 from solutions containing 0.5 M NaCl. Modification of porous membranes to bind proteins will most likely involve adsorption of only a few polyelectrolyte bilayers to simplify the process and avoid plugging of pores.  Moreover, the films should contain metal-ion complexes for capture of proteins through metal-ion affinity interactions (Figure 2.1).  Thus, we also examined swelling of (PAH/CMPEI)2 and (PAH/PDCMAA)2 films containing Cu2+ complexes.  These studies employed binding buffer 1 (pH 6.0) to match subsequent Con A-binding studies, as Con A solutions are not stable at pH 7.4. Figure 2.10 shows that for all film-adsorption pH values (pH 2 to 9), the (PAH/CMPEI)2-Cu2+ swelling in pH 6.0 buffer is around 200%. In pH 7.4 buffer the swelling     56 of a (PAH/CMPEI)2-Cu2+ film (deposited at pH 3 with 0.5 M NaCl) is still only 220%.  Thus, formation of the metal-ion complexes decreases film swelling, probably because Cu2+-iminodiacetate complexes have no net charge.  When immersed in pH 6.0 buffer, the (PAH/PDCMAA)2-Cu2+ films show average swellings of only 100% for deposition pH vales of 3, 5, or 7. Although both CMPEI and PDCMAA contain iminodiacetate moieties, the amine or ammonium groups in the backbone of CMPEI films likely increase swelling compared to films with PDCMAA, which contains a hydrocarbon backbone.   . Swelling of (PAH/CMPEI)2-Cu2+ (blue diamonds) and (PAH/PDCMAA)2-Cu2+ (red squares) films as a function of their deposition pH.  Films were deposited on MPA-modified, Au-coated Si wafers from polyelectrolyte solutions containing 0.5 M NaCl. All the swelling tests were performed in binding buffer 1 (pH 6).   Initial studies of protein binding examined capture of Con A in (PAH/CMPEI)2-Cu2+ and (PAH/PDCMAA)2-Cu2+ films adsorbed on MPA-modified Au-coated Si wafers. Binding presumably occurs when histidine groups on the protein coordinate with immobilized Cu2+.  050100150200250300350pH 2pH 3pH 5pH 7pH 9Swelling (%)Deposition pH  57 Using reflectance FTIR spectroscopy, we determine the amount of protein binding based on the amide absorbance, which we compare to the absorbance in spin-coated films with different thicknesses.32  (PAH/PDCMAA)2-Cu2+ films have average thicknesses ranging from 7-25 nm, depending on the deposition pH (see Figure 2.11), but these coatings bind the equivalent of <3 nm of protein, or less than a monolayer. (The dimensions of a Con A protomer, Mw=25,500 Da,  are 4.24.0 3.9 nm.43)  Even with an extra bilayer, (PAH/PDCMAA)3-Cu2+ films with a thickness of ~60 nm (deposited at pH 2) bind only 8 nm of Con A. Such limited binding will lead to low capacities in membranes modified with these films. In contrast, (PAH/CMPEI)2-Cu2+ films adsorbed at pH 2 have an average thickness around 40 nm and capture 18 nm of protein (Figure 2.11).  Adsorption of (PAH/CMPEI)2 at deposition pH values from 3-7 leads to thinner films than adsorption at pH 2 and binding of 5 nm of protein (Figure 2.11).  Thus, polyelectrolyte adsorption at low pH to obtain relatively thick CMPEI films and high swelling is likely vital to achieving high binding capacities.   . Thicknesses of (PAH/PDCMAA)2 and (PAH/CMPEI)2 multilayers after complexation of Cu2+, and the equivalent thicknesses of Con A subsequently adsorbed in these films.  PEMs were deposited from polyelectrolyte solutions containing 0.5 M NaCl at various pH values.  051015202530354045pH 2pH 3pH 5pH 7pH 9Thickness (nm)Deposition pH(PAH/CMPEI)2-Cu2+(PAH/PDCMAA)2-Cu2+ Con A binding(PAH/CMPEI)2-Cu2+ Con A binding(PAH/PDCMAA)2-Cu2+ 58  Adsorption of (PAH/CMPEI)n and (PAH/PDCMAA)n films within membrane pores is difficult to quantify.  To qualitatively assess the amount of adsorbed polymer we examined Cu2+ and Ni2+ binding in membranes modified with polyelectrolyte films.  As Figure 2.12 shows, an untreated nylon membrane modified with PAH/CMPEI (far left data bars) binds <1 mg of Cu2+ per mL of membrane.   This implies minimal adsorption of PAH/CMPEI, so we treated the nylon substrates with 0.1 M sodium chloroacetate in 3 M NaOH to increase the number of COOH groups on pore surfaces and enhance polyelectrolyte adsorption. Unfortunately, in control experiments carboxymethylated (CM) nylon captures 3 mg of Cu2+ per mL of membrane.  However, adsorption of PAH in the membrane decreases the Cu2+ capture to about 2 mg/mL, presumably because PAH forms salt bridges with some COO- groups to prevent binding.  Protonation of the amine groups should prevent them from binding Cu2+. (The pH of the Cu2+ loading solution is ~4).  Subsequent adsorption of a CMPEI layer leads to capture of 7 mg of Cu2+ per mL of membrane, and CM nylon membranes modified with single PAH/CMPEI and PAH/PDCMAA bilayers show similar Cu2+ binding. Moreover, (PAH/CMPEI)2- and (PAH/PDCMAA)2-modified CM membranes capture around 12 and 14 mg of Cu2+ per mL of membrane, respectively.  59  . Cu2+ (red bars) and Ni2+ (blue bars) binding capacities in PAH/CMPEI-modified nylon, carboxymethylated (CM) nylon, PAH-modified CM nylon, PAH/CMPEI-modified CM nylon, PAH/PDCMAA-modified CM nylon, (PAH/CMPEI)2-modified CM nylon, and (PAH/PDCMAA)2-modified CM nylon membranes. All polyelectrolytes were adsorbed at pH 2 from solutions containing 0.5 M NaCl. Error bars are the differences between experiments with two different membranes.  Importantly, the PAH/CMPEI-CM nylon membrane binds 16 times the amount of Cu2+ captured in an untreated nylon membrane modified with PAH/CMPEI.  Figure 2.13 shows SEM images of bare nylon, CM nylon, (PAH/CMPEI)-Cu2+ CM nylon and (PAH/CMPEI)2-Cu2+ CM nylon. The structures of the nylon membranes show no obvious change after carboxymethylation, so the primary effect of this treatment is the formation of COOH groups that facilitate adsorption of the initial PAH layer.   024681012141618   60  . SEM images of (A) nylon, (B) carboxymethylated nylon, (C) PAH/CMPEI-Cu2+-modified carboxymethylated nylon and (D) (PAH/CMPEI)2-Cu2+-modified carboxymethylated nylon membranes.  The scale bar is common to all images.  Selective capture of His-tagged proteins typically employs immobilized Ni2+ or Co2+ complexes, not Cu2+. Histidine binding to Ni2+ and Co2+ is weaker than to Cu2+ and thus requires multiple histidine residues for protein capture, which affords selective sorption of His-tagged species. As Figure 2.12 shows, CM nylon membranes modified with PAH/CMPEI and (PAH/CMPEI)2 films bind 2 mg/mL and 5 mg/mL of Ni2+, respectively.  This is considerably less than the Cu2+ binding capacity, perhaps because Ni2+ only binds strongly to sites with the full iminodiacetic acid functionality. Amines modified with a single carboxylic acid group (see Scheme 2.1) may not give stable Ni2+ complexes. The unmodified CM nylon also shows less Ni2+ binding than Cu2+ binding, and CM membranes modified with only PAH show minimal Ni2+ capture. PDCMAA contains only IDA binding groups, so there is not a large difference between Ni2+ and Cu2+ binding to membranes with PAH/PDCMAA films. Hence the       61 membranes modified with PAH/PDCMAA and (PAH/PDCMAA)2 capture more Ni2+ than corresponding membranes modified with PAH/CMPEI and (PAH/CMPEI)2. From metal-ion binding, we can estimate the polymer adsorption in a membrane. For (PAH/CMPEI)-modified membranes, the Ni2+ (Mw=58.7) binding is around 3 mg/mL. Assuming that only complete IDA groups bind Ni2+, 4 metal ions should bind to the CMPEI repeat unit in Scheme 1. Thus, a (PAH/CMPEI)-modified CM membrane will contain 14 mg/mL of CMPEI (repeat unit Mw =1112).   Due to the high cost of His-tagged proteins, we first employed Con A binding to Cu2+ complexes to evaluate the protein-binding capacities of membranes. Figure 2.14 shows the breakthrough curves for Con A capture in CM nylon membranes modified with PAH/CMPEI-Cu2+(purple circles) and PAH/PDCMAA-Cu2+ (green squares) films. Even though both films show similar Cu2+ binding (Figure 2.12), the total Con A bound to the membrane with PAH/CMPEI-Cu2+ is 59±5 mg/mL, whereas the membrane with PAH/PDCMAA-Cu2+ captures just 30±5 mg/mL.  Binding capacities determined from Con A elution with 50 mM EDTA are similar to those from the breakthrough curves (55±10 mg/mL and 35±8 mg/mL for PAH/CMPEI-Cu2+ and PAH/PDCMAA-Cu2+, respectively). The higher binding capacity with PAH/CMPEI-Cu2+ than PAH/PDCMAA-Cu2+ is consistent with the trends in Con A binding capacities of PEMs on Au-coated Si wafers (Figure 2.11).   62  . Breakthrough curves of Con A capture in CM nylon membranes (2.0-cm diameter) modified with PAH/CMPEI-Cu2+ (purple circles) and PAH/PDCMAA-Cu2+ (green squares). Both films were deposited at pH 2 with 0.5 M NaCl. The feed Con A concentration was 0.3 mg/mL and the volume flux was 10 cm/h.   . Breakthrough curves for Con A capture in (PAH/PDCMAA)2-Cu2+- (green squares) and (PAH/CMPEI)2-Cu2+- (purple circles) modified CM nylon membranes (2.0-cm diameter).  The films were deposited at pH 2 with 0.5 M NaCl in the polyelectrolyte solution, the feed Con A concentration was 0.3 mg/mL, and the volume flux was 10 cm/h. We also tested Con A binding in (PAH/CMPEI)2-Cu2+-modified CM nylon. Based on breakthrough curves (e.g., Figure 2.15), the Con A binding capacity in these membranes is 39±5 mg/mL, or less than in membranes with PAH/CMPEI-Cu2+ films. The unexpected 0.000.050.100.150.200.250.3002468101214161820Concentration  (mg/mL)Effluent volume (mL)00.050.10.150.20.250.302468101214161820Concentration (mg/mL)Effluent volume (mL) 63 decrease in binding compared to a film with a single bilayer may reflect decreased swelling with more bilayers or limited access to some small pores after coating the spongy membrane structure (see Figure 2.13) with two bilayers. Con A capture in membranes modified with (PAH/PDCMAA)2-Cu2+ is also less than in membranes with (PAH/PDCMAA)-Cu2+ (see Figure 2.15).  Because they showed the highest Con A capture, we determined the binding capacity for His-tagged ubiquitin using CM nylon membranes modified with PAH/CMPEI films.  However, in this case we employed the Ni2+ complex, which is necessary for selective capture of His-tagged protein.  Based on breakthrough curves (Figure 2.16), the binding capacity is ~60 mg/mL, and protein elution gave a capacity of ~70 mg/mL. This His-U binding is about 2/3 of what we previously obtained using polymer brush- or (PAA/PEI/PAA)-NTA-Ni2+-modified membranes (~90 mg/mL membrane).22,44 However, this new strategy avoids the challenges of growing polymer brushes or the expensive reaction of PAA/PEI/PAA with aminobutyl NTA. The dynamic binding capacity, i.e. the amount of protein bound when the effluent concentration is 10 % of the loading concentration, is around 30 mg/mL.  64  . Breakthrough curve for His-tagged ubiquitin capture in a (PAH/CMPEI)-modified CM membrane. The flow rate was 10 cm/h, the membrane had a diameter of 1.0-cm, and the feed His-tagged ubiquitin concentration was 0.3 mg/mL. The His-tagged ubiquitin binding capacity was 55 mg/mL for this membrane and 64 mg/mL for a second replicate membrane.  To demonstrate that membranes can isolate His-tagged protein directly from cell extracts, we purified His-tagged SUMO protein that was over-expressed in E. coli.  Figure 2.17 shows the SDS-PAGE analysis of a cell extract that contained His-tagged SUMO (lane 2), the same cell extract after passing through a (PAH/CMPEI)-modified CM membrane (lane 3), and the eluate (lane 4) from the membrane loaded with the cell extract. Notably, the effluent of the loading solution contains minimal His-tagged SUMO protein, and the only detectable band from the eluate stems from the His-tagged SUMO protein. Thus the membranes selectively capture His-tagged protein.   0.000.050.100.150.200.250.300.3502468Concentration (mg/mL)Effluent Volume (mL) 65  . SDS-PAGE analysis of purification of overexpressed His-tagged SUMO protein from an E. coli. lysate. Lane 1: molecular marker; Lane 2: cell lysate containing His-tagged SUMO protein; Lane 3: the cell lysate after passing through a (PAH/CMPEI)-Ni2+-modified CM membrane; Lane 4: the eluate of the loaded membrane.    Low metal-ion leaching is sometimes important to avoid contaminating protein solutions. Thus, we examined leaching from several modified membranes and a common commercial Ni2+ column. Membranes modified with one and two bilayers of PAH/CMPEI-Ni2+ or PAH/PDCMAA-Ni2+ (deposited at pH 2 in 0.5 M NaCl) were washed with 5 mL each (160 bed volumes) of binding buffer 2, washing buffer 1, washing buffer 2, stripping buffer, and 2% HNO3. (We summed the amounts of Ni2+ in the stripping buffer and HNO3.) The GE HitrapTM FF Ni column with a 1-mL bed volume was washed with 160 mL (160 bed volumes) each of binding buffer 2 and washing buffers 1 and 2. Subsequently, the remaining Ni2+ was eluted from the column with 15 mL of stripping buffer (elution was complete with EDTA so 2%        His-tagged SUMO  66 HNO3 was not needed). All the solutions were analyzed by atomic absorption spectroscopy.  . Ni2+ leaching from a GE HitrapTM FF Ni column and CM nylon membranes modified with (PAH/PDCMAA), (PAH/PDCMAA)2, (PAH/CMPEI) and (PAH/CMPEI)2 films. The numbers represent the percentage of Ni2+ ion lost in each solution. All the substrates were treated with 160 bed volumes (each) of binding buffer 2, washing buffers 1 and 2, elution buffer, and stripping buffer. The experiment was repeated twice for all substrates, and errors are differences between two trials.   Table 2.2 shows the leaching from the GE HitrapTM FF Ni column and different membranes as a percentage of the total Ni2+ binding. The (PAH/CMPEI)- and (PAH/CMPEI)2-modified membranes show the least percentage leaching in the binding and washing buffers, and the percentage of leaching in the elution buffer is within a factor of ~2 for all systems, although the GE column shows the lowest leaching in that buffer. The low leaching in the elution buffer for the GE column partly reflects the high leaching in the binding buffer. For all systems, the higher leaching in the elution buffer (0.5 M imidazole) than in the washing buffers stems from the formation of imidazole-Ni2+ complexes. Nevertheless, all the membrane substrates had less than 10 ppm Ni2+ in the 5 mL of elution buffer except the membrane modified with  67 (PAH/PDCMAA)2, which had 12.9 ± 1.1 ppm Ni2+. (Note the values in Table 2.2 are percentages of the total Ni2+ loaded and not concentrations.) The Ni2+ binding capacity of the GE HitrapTM FF Ni column is 1.6 ± 0.2 mg/mL, and Figure 2.12 shows that the Ni2+ binding capacities for all the membranes are higher than that for the Ni column. (For example, the Ni2+ binding capacity of the (PAH/CMPEI)-modified membrane is 2.7 mg/mL.) Overall, the metal leaching from all the substrates is similar, which is not surprising given that they likely have related ligands.  Adsorption of (PAH/CMPEI)-Ni2+ films may prove sufficiently simple and inexpensive to provide disposable, functional membranes. However, membrane reuse is always desirable, and the high swelling of PAH/CMPEI films (as much as 680%, see section 2.2) in buffer may lead to partial polyelectrolyte desorption. We evaluated the stability of CMPEI-containing films both on wafers and in membranes. For (PAH/CMPEI)2 films on Au-coated Si wafers (deposited on a MPA monolayer at pH 2 in 0.5 M NaCl), immersion for 20 h in binding buffer 2 (pH 7.4) led to only a 10% decrease in thickness, most of which occurred in the first 4 h (see Figure 2.18). Absorbances in reflectance IR spectroscopy also decreased about 10%, suggesting that the change in thickness results from a small loss of film and not simply deswelling or a change in conformation.   68  . Ellipsometric thicknesses of (PAH/CMPEI)2 films on MPA-modified, Au-coated Si wafers after immersion in binding buffer 2 for different times. The film was deposited at pH 2 from solutions containing 0.5 M NaCl. Error bars represent the standard deviations of measurements on at least three different films.  Films were rinsed with water prior to determining their thickness.   . UV/Vis absorbance at 595 nm and the concentration of CMPEI in the effluent binding buffer 2 passing through a (PAH/CMPEI)-Ni2+-effluent was added to 1.5 mL of Bradford assay dye for UV/Vis analysis. The concentration of CMPEI was determined by TOC using CMPEI solutions (0-10 ppm) as standards.  051015202502468101214161820Film thickness (nm)Time (h)01234500.010.020.030.04020406080UV/Vis absorbance at 585 nmEffluent volume (mL)Concentration of CMPEI (ppm)   69 Using TOC analysis, we determined the amount of the polyelectrolyte film lost during passage of binding buffer 2 (pH 7.4) through a membrane. After forming a (PAH/CMPEI) film and rinsing with only water, the first 20 mL of washing buffer passed through the membrane contained around 4 ppm of polymer (we assumed that the leaching was only due to CMPEI and used 1-10 ppm CMPEI solutions as standards). This corresponds to <20% of the total polymer based on our estimate of 14 mg of CMPEI/mL of membrane (the membrane volume in these leaching studies was 0.035 cm3, diameter 2-cm). Subsequent buffer washes contained <0.005 ppm (TOC detection limit) of polymer. Additionally, we added wash solutions to the Bradford dye and tested the absorbance at 595 nm (Figure 2.19) as in a typical Bradford assay.  The first mL of washing solution gave an absorbance of 0.02, which is equivalent to the absorbance given by 0.03 mg/mL of Con A. This absorbance rapidly declines and was only 0.002 after passing 20 mL of washing buffer through the membrane. In a typical protein-binding test, we wash the membranes with 40 mL of binding buffer prior to loading protein.  However, some breakthrough curves such as that for (PAH/CMPEI)2-Cu2+ (Figure 2.15) show a small and decreasing Bradford assay signal over the first 1-2 mL of protein loading. This may indicate that protein replaces a small amount of polyelectrolyte, i.e. the initial loading solution might contain 5 ppm of polyelectrolyte after passing through the membrane. We did not see this issue in binding of His U. As a further test of membrane stability, we performed 4 cycles of loading and elution of Con A in (PAH/CMPEI)-Cu2+-modified CM membranes. The Con A binding decreased by 40% (from 58 mg/mL to 35 mg/mL) over four cycles of loading, recharging with Cu2+, and elution (Figure 2.20). Thus, reuse is possible, but performance declines with use.  70  . Con A binding capacities of (PAH/CMPEI)-Cu2+-modified CM nylon membranes (blue bars and red bars represent two different membranes) in repeated measurements. Con A (0.3 mg/mL in binding buffer 1) was loaded and eluted four times, and membranes were recharged with Cu2+ before each capture experiment. The error bars are the differences between Con A binding capacities determined from the breakthrough curve and elution.   This study presents a facile method, LbL adsorption of functional polyelectrolytes, to modify membranes with metal-ion complexes that selectively capture His-tagged proteins. PAH/CMPEI adsorption yields a membrane with a His-tagged ubiquitin binding capacity of ~60 mg/mL, which is equal to the capacity of high-binding commercial beads. Moreover, these (PAH/CMPEI)-modified membranes show less than 10 ppm of Ni2+ in the elution buffer (0.5 M imidazole). Membranes modified with PAH/CMPEI show about twice the protein binding of corresponding membranes modified with PAH/PDCMAA, presumably because of more swelling with PAH/CMPEI. The His-tagged protein-binding capacity of the (PAH/PEI)-Ni2+-modified membranes is 2/3 of that for membranes modified through growth of polymer brushes or LbL adsorption of PAA/PEI/PAA followed by derivatization. However, direct adsorption of 0102030405060701234Con A binding capacity (mg/mL)Replicate 71 PAH and CMPEI in membranes is much simpler and less expensive than previous membrane modification methods and may lead to inexpensive, disposable membranes for rapid purification of His-tagged protein.                72                       73     (1) Lichty, J. J.; Malecki, J. L.; Agnew, H. D.; Michelson-Horowitz, D. J.; Tan, S., Comparison of Affinity Tags for Protein Purification, Protein Expr. Purif. 2005, 41, 98-105.   (2) Porath, J., Immobilized Metal-Ion Affinity-Chromatography, Protein Expr. Purif. 1992, 3, 263-281.   (3) Arnau, J.; Lauritzen, C.; Petersen, G. E.; Pedersen, J., Current Strategies for the Use of Affinity Tags and Tag Removal for the Purification of Recombinant Proteins, Protein Expr. Purif. 2006, 48, 1-13.   (4) Kawai, T.; Saito, K.; Lee, W., Protein Binding to Polymer Brush, Based on Ion-Exchange, Hydrophobic, and Affinity Interactions, J. Chromatogr. B 2003, 790, 131-142.   (5) Bhut, B. V.; Husson, S. M., Dramatic Performance Improvement of Weak Anion-Exchange Membranes for Chromatographic Bioseparations, J. Membr. Sci. 2009, 337, 215-223.   (6) Ghosh, R., Protein Separation Using Membrane Chromatography: Opportunities and Challenges, J. Chromatogr. A 2002, 952, 13-27.   (7) Datta, S.; Bhattacharyya, D.; Ray, P. D.; Nath, A.; Toborek, M., Effect of Pre-Filtration on Selective Isolation of Tat Protein by Affinity Membrane Separation: Analysis of Flux, Separation Efficiency, And Processing Time, Sep. Sci. Technol. 2007, 42, 2451-2471.   (8) Wang, J.; Sproul, R. T.; Anderson, L. S.; Husson, S. M., Development of Multimodal Membrane Adsorbers for Antibody Purification Using Atom Transfer Radical Polymerization, Polymer 2014, 55, 1404-1411.   (9) Yin, D. X.; Ulbricht, M., Antibody-Imprinted Membrane Adsorber via Two-Step Surface Grafting, Biomacromolecules 2013, 14, 4489-4496.   (10) Thommes, J.; Etzel, M., Alternatives to Chromatographic Separations, Biotechnol. Prog. 2007, 23, 42-45.   (11) Roper, D. K.; Lightfoot, E. N., Separation of Biomolecules Using Adsorptive Membranes, J. Chromatogr. A 1995, 702, 3-26.   (12) Thommes, J.; Kula, M. R., Membrane Chromatography - An Integrative Concept in the Downstream Processing of Proteins, Biotechnol. Prog. 1995, 11, 357-367.   (13) Saxena, A.; Tripathi, B. P.; Kumar, M.; Shahi, V. K., Membrane-Based Techniques  74 for the Separation and Purification of Proteins: An Overview, Adv. Colloid Interface Sci. 2009, 145, 1-22.   (14) Zeng, X. F.; Ruckenstein, E., Membrane chromatography: Preparation and Applications to Protein Separation, Biotechnol. Prog. 1999, 15, 1003-1019.   (15) Sun, L.; Dai, J. H.; Baker, G. L.; Bruening, M. L., High-Capacity, Protein-Binding Membranes Based on Polymer Brushes Grown in Porous Substrates, Chem. Mat. 2006, 18, 4033-4039.   (16) Bhut, B. V.; Conrad, K. A.; Husson, S. M., Preparation of High-Performance Membrane Adsorbers by Surface-Initiated AGET ATRP in the Presence of Dissolved Oxygen and Low Catalyst Concentration, J. Membr. Sci. 2012, 390, 43-47.   (17) Yang, Q.; Ulbricht, M., Cylindrical Membrane Pores with Well-Defined Grafted Linear and Comblike Glycopolymer Layers for Lectin Binding, Macromolecules 2011, 44, 1303-1310.   (18) Kawakita, H.; Masunaga, H.; Nomura, K.; Uezu, K.; Akiba, I.; Tsuneda, S., Adsorption of Bovine Serum Albumin to A Polymer Brush Prepared by Atom-Transfer Radical Polymerization in A Porous Inorganic Membrane, J. Porous Mat. 2007, 14, 387-391.   (19) Honjo, T.; Hoe, K.; Tabayashi, S.; Tanaka, T.; Shimada, J.; Goto, M.; Matsuyama, H.; Maruyama, T., Preparation of Affinity Membranes Using Thermally Induced Phase Separation for One-Step Purification of Recombinant Proteins, Anal. Biochem. 2013, 434, 269-274.   (20) Bhut, B. V.; Wickramasinghe, S. R.; Husson, S. M., Preparation of High-Capacity, Weak Anion-Exchange Membranes for Protein Separations Using Surface-Initiated Atom Transfer Radical Polymerization, J. Membr. Sci. 2008, 325, 176-183.   (21) Ulbricht, M.; Yang, H., Porous polypropylene membranes with different carboxyl polymer brush layers for reversible protein binding via surface-initiated graft copolymerization, Chem. Mat. 2005, 17, 2622-2631.   (22) Bhattacharjee, S.; Dong, J. L.; Ma, Y. D.; Hovde, S.; Geiger, J. H.; Baker, G. L.; Bruening, M. L., Formation of High-Capacity Protein-Adsorbing Membranes through Simple Adsorption of Poly(acrylic acid)-Containing Films at Low pH, Langmuir 2012, 28, 6885-6892.   (23) Naka, K.; Tachiyama, Y.; Hagihara, K.; Tanaka, Y.; Yoshimoto, M.; Ohki, A.; Maeda, S., Synthesis and Chelating Properties of Poly (N,N-dicarboxymethyl)allylamine Derived from Poly(allylamine), Polym. Bull. 1995, 35, 659-663.   (24) Sheng, C. J.; Wijeratne, S.; Cheng, C.; Baker, G. L.; Bruening, M. L., Facilitated Ion Transport Through Polyelectrolyte Multilayer Films Containing Metal-Binding Ligands, J.  75 Membr. Sci. 2014, 459, 169-176.   (25) Wijeratne, S.; Bruening, M. L.; Baker, G. L., Layer-by-Layer Assembly of Thick, Cu2+-Chelating Films, Langmuir 2013, 29, 12720-12729.   (26) Anzai, J.; Kobayashi, Y.; Nakamura, N.; Nishimura, M.; Hoshi, T., Layer-by-Layer Construction of Multilayer Thin Films Composed of Avidin and Biotin-Labeled Poly(amine)s, Langmuir 1999, 15, 221-226.   (27) Qiagen, http://www.qiagen.com/us/products/catalog/sample-technologies/protein-sample-technologies/purification-kits-and-resins/ni-nta-agarose#technicalspecification, 08/06/2016.    (28) Clonetech, His60 Nickel Resin Beats Ni-NTA Resin for His-Tag Purification, http://www.clontech.com/US/Support/Applications/Tagged_Protein_Purification/Ni-NTA_Resin_vs_His60, 08/06/2016.    (29) Bucur, C. B.; Sui, Z.; Schlenoff, J. B., Ideal Mixing in Polyelectrolyte Complexes and Multilayers: Entropy Driven Assembly, J. Am. Chem. Soc. 2006, 128, 13690-13691.   (30) Vig, J. R., Uv Ozone Cleaning Of Surfaces, J. Vac. Sci. Technol. A 1985, 3, 1027-1034.   (31) Miller, M. D.; Bruening, M. L., Correlation of the Swelling and Permeability of Polyelectrolyte Multilayer Films, Chem. Mat. 2005, 17, 5375-5381.   (32) Dai, J. H.; Bao, Z. Y.; Sun, L.; Hong, S. U.; Baker, G. L.; Bruening, M. L., High-Capacity Binding of Proteins By Poly(Acrylic Acid) Brushes and Their Derivatives, Langmuir 2006, 22, 4274-4281.   (33) Nakamoto, K.; Morimoto, Y.; Martell, A. E., Infrared Spectra of Aqueous Solutions .2. Iminodiacetic Acid, N-Hydroxyethyliminodiacetic Acid and Nitrilotriacetic Acid, J. Am. Chem. Soc. 1962, 84, 2081-&.   (34) Pack, D. W.; Arnold, F. H., Langmuir Monolayer Characterization of Metal Chelating Lipids for Protein Targeting to Membranes, Chem. Phys. Lipids 1997, 86, 135-152.   (35) Hoffmann, K.; Tieke, B., Layer-by-Layer Assembled Membranes Containing Hexacyclen-Hexaacetic Acid And Polyethyleneimine N-Acetic Acid and Their Ion Selective Permeation Behaviour, J. Membr. Sci. 2009, 341, 261-267.   (36) McAloney, R. A.; Sinyor, M.; Dudnik, V.; Goh, M. C., Atomic Force Microscopy Studies of Salt Effects on Polyelectrolyte Multilayer Film Morphology, Langmuir 2001, 17, 6655-6663.  76   (37) Schlenoff, J. B.; Dubas, S. T., Mechanism of Polyelectrolyte Multilayer Growth: Charge Overcompensation and Distribution, Macromolecules 2001, 34, 592-598.   (38) Dobrynin, A. V.; Rubinstein, M.; Joanny, J. F., Adsorption of A Polyampholyte Chain on a Charged Surface, Macromolecules 1997, 30, 4332-4341.   (39) Jeon, J.; Dobrynin, A. V., Monte Carlo Simulations of Polyampholyte-Polyelectrolyte Complexes: Effect of Charge Sequence and Strength of Electrostatic Interactions, Phys. Rev. E 2003, 67.   (40) Bowman, W. A.; Rubinstein, M.; Tan, J. S., Polyelectrolyte-Gelatin Complexation: Light-Scattering Study, Macromolecules 1997, 30, 3262-3270.   (41) Fu, J. H.; Ji, J.; Shen, L. Y.; Kueller, A.; Rosenhahn, A.; Shen, J. C.; Grunze, M., pH-Amplified Exponential Growth Multilayers: A Facile Method to Develop Hierarchical Micro- and Nanostructured Surfaces, Langmuir 2009, 25, 672-675.   (42) Bieker, P.; Schonhoff, M., Linear and Exponential Growth Regimes of Multi layers of Weak Polyelectrolytes in Dependence on pH, Macromolecules 2010, 43, 5052-5059.   (43) Becker, J. W.; Reeke, G. N.; Wang, J. L.; Cunningham, B. A.; Edelman, G. M., Covalent and 3-Dimensional Structure of Concanavalin-A .3. Structure Of Monomer and Its Interactions with Metals and Saccharides, J. Biol. Chem. 1975, 250, 1513-1524.   (44) Anuraj, N.; Bhattacharjee, S.; Geiger, J. H.; Baker, G. L.; Bruening, M. L., An All-Aqueous Route to Polymer Brush-Modified Membranes with Remarkable Permeabilites and Protein Capture Rates, J. Membr. Sci. 2012, 389, 117-125.         77 Chapter 3. Enzymatic Membrane Reactor for Affinity Tag Removal     78     79      80      81      82             83            84      85      86        87    88      89     y = 0.63xR² = 0.9900.10.20.30.40.50.60.700.20.40.60.811.2AbsorbanceConcentration mg/mL 90       91    R² = 0.99R² = 0.9827R² = 0.9907024681012141618020000400006000080000100000120000of His-tagged SUMO branching enzymeIntensity 92            93       94          95     96            97                        98                                                99                     100 Chapter 4. Controlled Proteolysis in Porous Membrane Reactors Containing Immobilized Trypsin       101   102  -caseindenatured bovine serum albumin. Increased flow rates through the membrane (shorter residence times)    and nylon (0.45 µm, Millipore, HNWP02500 or 5.0 ) membranes were employed as substrates for trypsin immobilization. Trypsin from bovine pancreas (type I, 12200 units/mg solid), benzamidine hydrochloride, N-benzoyl-L-arginine ethyl ester hydrochloride (BAEE), poly(acrylic acid) (PAA, average molecular weight ~100,000 Da, 35%  103 aqueous solution), poly(sodium 4-styrenesufonate) (PSS, Mw ~ 70,000 Da), N-(3-dimethylaminopropyl)--ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide -casein, bovine heart cytochrome c, apomyoglobin (protein sequencing grade from horse skeletal muscle, salt-free lyophilized powder), and myoglobin (from horse skeletal muscle 95-100%, salt-free lyophilized powder) were purchased from Sigma-Aldrich.  Buffers were prepared using analytical grade chemicals and deionized (Milli-     104      105     106       107            N-(3-dimethylaminopropyl)--ethylcarbodiimide hydrochloride;  108    109     110          (4.2)  111    , which is consistent with the typical        (4.4)  112    024681012141314151617181920Residence time (s)BAEE concentration (mM) 113             0501001502002500.20.40.60.81Time (s)BAEE (mM) 114     115      116     117    118     119       120      121     122    123        124    125        126             127                          128                                                    129                                            130                                            131                                                132                     133 Chapter 5. Rapid Protein Digestion and Purification with Membranes Attached to Pipette Tips    The previous chapters described the development of membranes for purification of His-tagged protein and controlled proteolysis. This chapter develops these membranes into a potential high-throughput platform by attaching them to the ends of pipette tips (Scheme 5.1).  When combined with automatic pipets, this platform enables proteolysis within 30 s and protein isolation within 2 minutes.  . Attaching functional membranes to the end of pipette tips for fast proteolysis and protein purification.   134     135      136       137      138    139         140     P        141    Protein Binding Enzyme 5.0 µm 1.2 µm Trypsin 6±1 mg/mL 11±1 mg/mL Pepsin 32±6 mg/mL 48±5 mg/mL   142     143     144     145    0204060801001234567891011121314151617181920Relative Intensity (100%) 146    147     148    149     150      151     152     153      154            155        156                         157                                                    158                                                  159                                             160 Chapter 6. Summary and future work   poly[(N,N-dicarboxymethyl) allylamine] (  161    162   direct adsorption of relatively inexpensive polyelectrolytes containing chelating groups effectively modifies membranes to bind metal ions and capture His-tagged protein. In addition to purification of His-tagged protein,   163       164     165       y = 105.47x -37.724R² = 0.992050010001500200025000.05.010.015.020.0Florescence intenistyConcentration (05001000150020002500300320340360380400Florescence IntensityWavelength (nm) 166     00.020.040.060.080.105101520Concentration (mg/mL)Effluent Volume (mL) 167      00.020.040.060.080.100.511.522.53Concentration (mg/mL)Effluent Volume (mL) 168    169         170                                              171                                                172