Alginate encapsulant incorporating CXCL12 supports long-term allo- and xenoislet transplantation without systemic immune suppression.

Islet transplantation represents a potentially curative approach for individuals with Type I Diabetes. The requirement for systemic immune suppression to control immune-mediated rejection of transplanted islets and the limited human islet supply represent significant roadblocks to progress for this approach. Islet microencapsulation in alginate offers limited protection in the absence of systemic immunosuppression, but does not support long-term islet survival. The chemokine, CXCL12, can repel effector T cells while recruiting immune-suppressive regulatory T cells (Tregs) to an anatomic site while providing a prosurvival signal for beta-cells. We proposed that coating or encapsulating donor islets with CXCL12 would induce local immune-isolation and protect and support the function of an allo- or xenograft without systemic immune suppression. This study investigated the effect of alginate microcapsules incorporating CXCL12 on islet function. Islet transplantation was performed in murine models of insulin-dependent diabetes. Coating of islets with CXCL12 or microencapsulation of islets with alginate incorporating the chemokine, resulted in long-term allo- and xenoislet survival and function, as well as a selective increase in intragraft Tregs. These data support the use of CXCL12 as a coating or a component of an alginate encapsulant to induce sustained local immune-isolation for allo- or xenoislet transplantation without systemic immunosuppression.

Microencapsulated islet allografts in diabetic NOD mice and nonhuman primates.

OBJECTIVE: Our goal was to assess the efficacy of encapsulated allogeneic islets transplanted in diabetic NOD mice and streptozotocin (STZ)-diabetic nonhuman primates (NHPs). MATERIALS AND METHODS: Murine or NHP islets were microencapsulated and transplanted in non-immunosuppressed mice or NHPs given clinically-acceptable immunosuppressive regimens, respectively. Two NHPs were treated with autologous mesenchymal stem cells (MSCs) and peri-transplant oxygen therapy. Different transplant sites (intraperitoneal [i.p.], omental pouch, omental surface, and bursa omentalis) were tested in separate NHPs. Graft function was monitored by exogenous insulin requirements, fasting blood glucose levels, glucose tolerance tests, percent hemoglobin A1c (% HbA1c), and C-peptide levels. In vitro assessment of grafts included histology, immunohistochemistry, and viability staining; host immune responses were characterized by flow cytometry and cytokine/chemokine multiplex ELISAS. RESULTS: Microencapsulated islet allografts functioned long-term i.p. in diabetic NOD mice without immunosuppression, but for a relatively short time in immunosuppressed NHPs. In the NHPs, encapsulated allo-islets initially reduced hyperglycemia, decreased exogenous insulin requirements, elevated C-peptide levels, and lowered % HbA1c in plasma, but graft function diminished with time, regardless of transplant site. At necropsy, microcapsules were intact and non-fibrotic, but many islets exhibited volume loss, central necrosis and endogenous markers of hypoxia. Animals receiving supplemental oxygen and autologous MSCs showed improved graft function for a longer post-transplant period. In diabetic NHPs and mice, cell-free microcapsules did not elicit a fibrotic response. CONCLUSIONS: The evidence suggested that hypoxia was a major factor for damage to encapsulated islets in vivo. To achieve long-term function, new approaches must be developed to increase the oxygen supply to microencapsulated islets and/or identify donor insulin-secreting cells which can tolerate hypoxia.

Cryoprotectant transport through articular cartilage for long-term storage: experimental and modeling studies.

OBJECTIVE: Long-term storage of articular cartilage (AC) remains challenging due to poor post-thaw viability. An initial step towards addressing this issue is characterizing cryoprotectant (CPA) transport, since ensuring adequate CPA equilibration throughout the tissue offers protection during cooling. This study takes a systematic approach in determining CPA transport rates through bovine AC and uses that information in mathematical models to determine CPA equilibration times. DESIGN: Diffusion of high concentration single (6.9 M dimethyl sulfoxide (DMSO)) and multi-component CPA solutions (VS55, 3.1 M DMSO+2.2 M 1,2-propanediol (PD)+3.1 M formamide (FM)) was measured through AC using (1)H nuclear magnetic resonance (NMR) imaging and localized spectroscopy, respectively. Using experimentally calculated effective diffusivities, diffusion models describing CPA transport through the tissue matrix and across chondrocyte membranes were combined to design a CPA addition and removal scheme for a cartilage plug of clinically relevant dimensions. RESULTS: (1)H NMR imaging and localized spectroscopy experiments suggested that the permeation of CPAs through AC (5 mm diameter, 5-10 mm in thickness) took on the order of 4 h for full equilibration at 22 degrees C. Imaging clearly showed the permeation of DMSO into cartilage over time and localized spectroscopy was able to distinguish the permeation rates of the individual VS55 components and water. Experimentally measured diffusivity values were used in CPA addition/removal simulations with a cartilage plug of clinically relevant dimensions (5 mm diameter, 2 mm in thickness). Results suggested a multi-step approach for adding and removing high concentration CPAs, with the addition and removal each taking approximately 2 h to complete. CONCLUSIONS: This study provides a foundation for designing CPA addition and removal protocols for effective long-term storage of cartilage tissue using a novel approach to measure CPA permeation.

Efficient transmicrocatheter delivery of functional fibroblasts with a bioengineered collagen gel-platinum microcoil complex: toward the development of endovascular cell therapy for cerebral aneurysms.

BACKGROUND AND PURPOSE: Endoaneurysmal implantation of fibroblasts may promote healing of aneurysms and reduce recanalization after therapeutic embolization. The purpose of our study was to develop a device for delivery of fibroblasts with use of current microcoil technology. MATERIALS AND METHODS: Cell carrier devices and cell-free devices were fabricated by associating collagen gels (with or without fibroblasts) with platinum microcoils. During the propagation of control cell carrier devices for 1 week in culture, cell-mediated gel contraction (CMGC) occurred. Modified cell carrier devices created by glutaraldehyde cross-linking, ascorbate coculture, or extended CMGC were also characterized in vitro. Devices were deployed through microcatheters (533 microm lumen, 160 cm length). Gel retention, cell retention, cell death, and the ability to support local cell migration were analyzed in vitro. RESULTS: Cell viability was reduced by glutaraldehyde cross-linking but not by microcatheter transit. During microcatheter transit, cell carrier devices liberated minimal particulate matter and cellular DNA. Liberated particulate matter was reduced by glutaraldehyde cross-linking (P < .05) and extended CMGC (P < .04). Only cell carrier devices treated with glutaraldehyde cross-linking did not exhibit cell migration after microcatheter transit. Passage of cell-free devices through microcatheters sheared off most of their collagen gel. CONCLUSION: Collagen gel-platinum microcoil complexes can mediate efficient transmicrocatheter delivery of viable, migration-capable fibroblasts. CMGC is a necessary component of the process of gel stabilization that enables successful microcatheter transit. Although extended CMGC and glutaraldehyde cross-linking enhance gel stabilization, glutaraldehyde cross-linking decreases cell viability and migratory potential.

Effects of oxygen on metabolic and secretory activities of beta TC3 cells.

We have investigated the rates of glucose consumption, lactate production and insulin secretion by mouse insulinoma beta TC3 cells exposed to high glucose and oxygen concentrations in the range of 132 mmHg (normoxia) to 0 mmHg (anoxia). The rates of glucose consumption and lactate production, and the yield of lactate on glucose were 6.4 +/- 0.2 nmol/h - 10(5) cells, 7.7 +/- 0.5 nmol/h - 10(5) cells, and 1.2 +/- 0.1 respectively, at oxygen concentrations between 132-25 mmHg. These values increased gradually as the oxygen concentration was reduced below 25 mmHg, reaching a maximum value of 12.8 +/- 0.4, 23.8 +/- 1.1, 1.9 +/- 0.1 respectively, at complete anoxia. Insulin secretion remained constant at 360 +/- 24 pmol/h - 10(8) cells at oxygen concentrations between 132-7 mmHg, but was inhibited at lower oxygen concentrations, dropping to 96 +/- 24 pmol/h - 10(8) cells at 0 mmHg. The rate of insulin secretion in the presence of high glucose under anoxia was significantly higher than the rate of basal secretion (28.2 +/- 3.0 pmol/h - 10(8) cells) at normoxia. The secretory properties of beta TC3 cells at low oxygen concentrations may have implications in the development of a diffusion-based bioartificial tissue constructs for the long-term treatment of Insulin Dependent Diabetes Mellitus.

Noninvasive measurement of viable cell number in tissue-engineered constructs in vitro, using 1H nuclear magnetic resonance spectroscopy.

Noninvasive monitoring of tissue-engineered constructs is of critical importance for accurate characterization of constructs and their remodeling in vitro and in vivo. This study investigated the utility of (1)H NMR spectroscopy to noninvasively quantify viable cell number in tissue-engineered substitutes in vitro. Agarose disk-shaped constructs containing betaTC3 cells were employed as the model tissue-engineered system. Two construct prototypes containing different initial cell numbers were monitored by localized, water-suppressed 1H NMR spectroscopy over the course of 13 days. (1)H NMR measurements of the total choline resonance at 3.2 ppm were compared with results from the traditional cell viability assay MTT and with insulin secretion rates. Results show a strong linear correlation between total choline and MTT (R (2) = 0.86), and between total choline and insulin secretion rate (R (2) = 0.90). Overall, this study found noninvasive measurement of total choline to be an accurate and nondestructive assay for monitoring viable betaTC3 cell numbers in tissue-engineered constructs. The applicability of this method to in vivo monitoring is also discussed.

In vivo noninvasive monitoring of a tissue engineered construct using 1H NMR spectroscopy.

Direct, noninvasive monitoring of tissue engineered substitutes containing live, functional cells would provide valuable information on dynamic changes that occur postimplantation. Such changes include remodeling both within the construct and at the interface of the implant with the surrounding host tissue, and may result in changes in the number of viable cells in the construct. This study investigated the use of 1H NMR spectroscopy in noninvasively monitoring the viable cell number within a tissue engineered construct in vivo. The construct consisted of mouse betaTC3 insulinomas in a disk-shaped agarose gel, surrounded by a cell-free agarose gel layer. Localized 1H NMR spectra were acquired from within implanted constructs, and the total choline resonance was measured. Critical issues that had to be addressed in accurately quantifying total choline from the implanted cells included avoiding signal from host tissue and correcting for interfering signal from diffusing solutes. In vivo NMR measurements were correlated with MTT assays and NMR measurements performed in vitro on explanted constructs. Total choline measurements accurately and noninvasively quantified viable betaTC3 cell numbers in vivo, in the range of 1 x 10(6) to more than 14 x 10(6) cells, and monitored changes in viable cell number that occurred in the same construct over time. This is the first study using NMR techniques to monitor viable cell numbers in an implanted tissue substitute. It established architectural characteristics that a construct should have to be amenable to NMR monitoring, and it set the foundation for future in vivo investigations with other tissue engineered implants.

Vitrification of tissue engineered pancreatic substitute.

Despite significant advances, some critical issues remain for the long-term storage of an engineered pancreas. In this study we employed a tissue engineered pancreatic substitute model-insulin-secreting betaTC3 cells entrapped in calcium alginate/poly-L-lysine/alginate beads-to demonstrate that a prototype vitrification method can prevent ice formation and maintain cell viability/function. The results showed that the structure of the frozen samples was distorted by ice crystals throughout the matrix. In marked contrast, the vitrified samples appeared to be free of ice. Morphologic studies demonstrated extensive fractures and vacuolation in frozen specimens while there were no fractures in vitrified TEPSs. Both vitrified and frozen constructs showed some vacuolization compared to the control samples. Frozen beads showed a significantly decreased viability compared to fresh controls and the VS55 group (P < .001). There was no significant difference between the vitrified and fresh samples. Vitrification using the VS55 protocol shows similar viability and secretion properties to the control group of fresh beads. Vitrification using the PEG 400 protocol resulted in slightly lower viability and secretion properties relative to the control group; conventional freezing resulted in even significantly lower viability and secretion properties. These results combine to demonstrate feasibility of vitrification as a storage method for a tissue engineered pancreas.

The effects of alginate composition on encapsulated betaTC3 cells.

The effects of alginate composition on the growth of murine insulinoma betaTC3 cells encapsulated in alginate/poly-L-lysine/alginate (APA) beads, and on the overall metabolic and secretory characteristics of the encapsulated cell system, were investigated for four different types of alginate. Two of the alginates used had a high guluronic acid content (73% in guluronic acid residues) with varying molecular weight, while the other two had a high mannuronic acid content (68% in mannuronic acid residues) with varying molecular weight. Each composition was tested using two different polymer concentrations. Our data show that betaTC3 cells encapsulated in alginates with a high guluronic acid content experienced a transient hindrance in their metabolic and secretory activity because of growth inhibition. Conversely, betaTC3 cells encapsulated in alginates with a high mannuronic acid content experienced a rapid increase in metabolic and secretory activity as a result of rapid cell growth. Our data also demonstrate that an increase in either molecular weight or concentration of high mannuronic acid alginates did not alter the behavior of the encapsulated betaTC3 cells. Conversely, an increase in molecular weight and concentration of high guluronic acid alginates prolonged the hindrance of glucose metabolism, insulin secretion and cell growth. These observations can be best interpreted by changes in the microstructure of the alginate matrix, i.e., interaction between the contiguous guluronic acid residues and the Ca2+ ions, as a result of the different compositions.

Effects of alginate composition on the metabolic, secretory, and growth characteristics of entrapped beta TC3 mouse insulinoma cells.

The effects of alginate composition on cell growth as well as the metabolic and secretory profile of transformed beta-cells entrapped in alginate/poly-L-lysine/alginate (APA) solid beads were investigated following entrapment of beta TC3 mouse insulinoma cells in alginate composed of either high mannuronic acid or high guluronic acid residues. Entrapped cultures were maintained in spinner flasks for 40-60 days. The pattern of cell growth and the overall rates of glucose consumption and insulin secretion were investigated. Cultures of beta TC3 cells entrapped in alginate composed predominantly of mannuronic acid units (77%) displayed a linear increase in the rates of glucose consumption and insulin secretion concomitant with an increase in cell population in the periphery of the beads. Conversely, cultures of beta TC3 cells entrapped in alginate composed predominantly of high guluronic acid units (69%) displayed a decrease in the rates of glucose consumption and insulin secretion during the first three weeks of culture, followed by a rapid recovery that surpassed the initial rates by day 40. This biphasic pattern was concomitant to a decrease in viable cells during the first three weeks as ascertained by histology, followed by an increase in cell proliferation. Cell growth in high guluronic acid alginate took place at random locations throughout the solid bead and not in the periphery, as was the case in high mannuronic acid alginate preparations. Possible reasons for these differences and the significance of these findings in the context of a bioartificial pancreas composed of APA entrapped transformed cells are discussed.

Tissue engineering: from biology to biological substitutes.

Tissue engineering is an emerging multidisciplinary and interdisciplinary field involving the development of bioartificial implants and/or the fostering of tissue remodeling with the purpose of repairing or enhancing tissue or organ function. Bioartificial constructs generally consist of cells and biomaterials, so tissue engineering draws from both cell and biomaterials science and technology. Successful applications require a thorough understanding of the environment experienced by cells in normal tissues and by cells in bioartificial devices before and after implantation. This paper reviews these topics, as well as the current status and future possibilities in the development of different bioartificial constructs, including bioartificial skin, cardiovascular implants, bioartificial pancreas, and encapsulated secretory cells. Issues that need to be addressed in the future are also discussed. These include, but are not limited to, the development of new cell lines and biomaterials, the evaluation of the optimal construct architecture, and the reproducible manufacture and preservation of bioartificial devices until ready for use.

Effect of oxygen tension and alginate encapsulation on restoration of the differentiated phenotype of passaged chondrocytes.

The implantation of laboratory-grown tissue offers a valuable alternative approach to the treatment of cartilage defects. Procuring sufficient cell numbers for such tissue-engineered cartilage is a major problem since amplification of chondrocytes in culture typically leads to loss of normal cell phenotype yielding cartilage of inferior quality. In an effort to overcome this problem, we endeavored to regain the differentiated phenotype of chondrocytes after extensive proliferation in monolayer culture by modulating cell morphology and oxygen tension towards the in vivo state. Passaged cells were encapsulated in alginate hydrogel in an effort to regain the more rounded shape characteristic of differentiated chondrocytes. These cultures were exposed to reduced (5%-i.e., physiological), or control (20%) oxygen tensions. Both alginate encapsulation and reduced oxygen tension significantly upregulated collagen II and aggrecan core protein expression (differentiation markers). In fact, after 4 weeks in alginate at 5% oxygen, differentiated gene expression was comparable to primary chondrocytes. Collagen I expression (dedifferentiation marker) decreased dramatically after alginate entrapment, while reduced oxygen tension had no effect. It is concluded that alginate encapsulation and reduced oxygen tension help restore key differentiated phenotypic markers of passaged chondrocytes. These findings have important implications for cartilage tissue engineering, since they enable the increase in differentiated cell numbers needed for the in vitro development of functional cartilaginous tissue suitable for implantation.

Non-Invasive monitoring of a bioartificial pancreas in vitro and in vivo.

Monitoring biochemical processes relevant to the function, survival, and longevity of tissue-engineered pancreatic constructs is important for the development of an optimum construct design as well as patient care management after implantation. In this report we demonstrate the ability of nuclear magnetic resonance (NMR) techniques to monitor aspects of intracellular metabolism, overall morphology, and distribution of a microencapsulation based bioartificial pancreas in vitro and in vivo.

Effects of short-term hypoxia on a transformed cell-based bioartificial pancreatic construct.

Hypoxia is an adverse condition that can jeopardize the function of a bioartificial pancreatic construct. In this study we have investigated the effects of short-term hypoxic exposure (up to 24 h) on the bioenergetic status, metabolism, and insulin secretion of perfused pancreatic constructs composed of alginate/poly-L-lysine/alginate (APA) encapsulated mouse insulinoma betaTC3 cells. The bioenergetic status of the encapsulated cells was monitored noninvasively with the aid of 31P NMR spectroscopy, while glucose, lactate, and insulin concentrations were measured with off-line assays from media samples removed from the perfusion loop. Our results demonstrate that in freshly prepared constructs insulin secretion was not affected by the hypoxic conditions, although intracellular ATP concentration decreased and glucose consumption increased. Alternatively, in constructs that were maintained in our perfusion system for at least 10 days, identical hypoxic conditions resulted in a decreased insulin secretion concomitant to a decreased intracellular ATP concentration and increased glucose consumption. These results suggest that the effects of hypoxia on a transformed cell-based pancreatic construct are not constant throughout the duration of an in vitro culture. The observed differences are attributed to the significant cell growth and rearrangement that occurs with time during an in vitro culture of the constructs.

Towards the development of a bioartificial pancreas: effects of poly-L-lysine on alginate beads with BTC3 cells.

A bioartificial tissue construct that consists of insulin-secreting cells entrapped in an alginate/poly-L-lysine (PLL) matrix offers a promising approach for the treatment of type I diabetes. Use of transformed cells has been proposed as a solution to the cell availability problem posed by islets. The growth characteristics of transformed cells in their sequestered environment and the effects of PLL on their metabolic and secretory activities have not yet been characterized. Our data demonstrate that mouse insulinoma beta TC3 cells proliferate while they are entrapped in both PLL-free and PLL-coated alginate beads. During this process, cell aggregates develop in the bead periphery, which increase in number and size with time. PLL is crucial for the long-term in vitro structural stability of beads, and it does not appear to affect the metabolic and secretory activities of entrapped beta TC3 cells. The implications of these findings in the development of a bioartificial pancreatic construct based on transformed cells are discussed.

In vitro monitoring of total choline levels in a bioartificial pancreas: (1)H NMR spectroscopic studies of the effects of oxygen level.

This investigation implements specifically designed solvent-suppressed adiabatic pulses whose properties make possible the long-term monitoring of (1)H NMR detectable metabolites from alginate/poly-l-lysine/alginate (APA)-encapsulated betaTC3 cells. Our encapsulated preparations were maintained in a perfusion bioreactor for periods exceeding 30 days. During this prolonged cultivation period, the cells were exposed to repetitive hypoxic episodes of 4 and 24 h. The ratio of the total choline signal (3.20 ppm) to the reference signal (observed at 0.94 ppm assigned to isoleucine, leucine, and valine) decreased by 8-10% for the 4-h and by 20-32% for the 24-h episodes and returned to its prehypoxic level upon reoxygenation. The decrease in the mean value of total choline to reference signal ratio for three 4-h and two 24-h episodes in two different cultures was highly significant (P<0.01). The rate of recovery by this ratio was slower than the rates of recovery by oxygen consumption, lactate production, or glucose consumption. A step-up in oxygen level led to a new, higher value for the total choline to reference ratio. From spectra of extracts at 400 MHz, it was determined that 63.6% of the total choline signal is due to intracellular phosphorylcholine. Therefore, it is inferred that the observed changes in total choline signal are linked to an oxygen level dependence of the intracellular phosphorylcholine. Several possible mechanisms in which oxygen may influence phosphorylcholine metabolism are suggested. In addition, the implications of these findings to the development of a noninvasive monitoring method for tissue-engineered constructs composed of encapsulated cells are discussed.

Engineering challenges in the development of an encapsulated cell system for treatment of type 1 diabetes.

Implantation of glucose-responsive, insulin-secreting cells is promising in providing a treatment for type I diabetes, which is more effective, less invasive, and potentially less costly than conventional insulin injections. However, in spite of promising results with animal studies, a clinical product or therapeutic procedure based on encapsulated cells does not yet exist. This is because a number of barriers remain to be addressed, which include a source of functional cells, a stable, biocompatible membrane offering immune protection to the implant, a construct architecture ensuring cell viability and construct function, and the engineering of immune acceptance of the construct post-implantation. This article reviews these barriers and the current state-of-the-art, with special emphasis on the engineering challenges involved, and discusses possible ways to tackle the complex problems currently preventing this approach from reaching clinical practice.

Tissue engineering of a bioartificial pancreas: modeling the cell environment and device function.

Cell-based implantable artificial tissues are most promising for the long-term treatment of endocrine diseases, such as diabetes. One type of a bioartificial pancreas device consists of calcium alginate microbeads containing insulin-secreting cells and is surrounded by a poly(L-lysine) (PLL) membrane. The membrane is semipermeable, allowing cellular nutrients and metabolites to diffuse through but excluding the antibodies and cytotoxic cells of the host, thus immunoprotecting the cells. The device can be modeled by writing the equations for diffusion of nutrients and metabolites through the polymer and for consumption of the former and production of the latter by the cells. In this paper, we describe the construction and analysis of such a model for alginate/PLL microbeads with insulin-secreting recombinant mouse pituitary AtT-20 and mouse insulinoma beta TC3 cells. Entrapped AtT-20 cells are a simplified model system, whereas microbeads with beta TC3 cells constitute a realistic artificial pancreatic device. Effective diffusivities of key compounds through the polymer with entrapped, inactivated AtT-20 spheroids were measured first. The kinetics of glucose and oxygen consumption and insulin secretion were modeled next, and the equations for diffusion and reaction were then combined to describe the entire system. The model was used to compute nutrient and metabolite concentration profiles in beads and the bead secretory response for different bead sizes and cell loadings. The size and loading necessary for the cells to be well nourished and for the beads to be rapidly responsive to step-ups and step-downs of secretion stimuli were evaluated. It was shown that if the cells are hypersensitive to glucose, i.e., they do not shut off secretion at the physiological glucose threshold but at a lower one, so are the microbeads. This work demonstrates the usefulness of mechanistic models with representative parameter values in optimizing the design of artificial tissues and in characterizing aspects of their behavior that are of importance for restoring in vivo function.

Quantitative modeling of limitations caused by diffusion.

Transport of nutrients and metabolites in many bioartificial tissue constructs relies exclusively on diffusion, i.e., on the presence of a concentration gradient between the inside of the construct and the surrounding milieu. A quantitative evaluation of the rate of diffusional processes is thus essential for properly designing three-dimensional cell-polymer systems, and for assessing the chemical environment at various locales within the construct.

Ammonium selectively inhibits the regulated pathway of protein secretion in two endocrine cell lines.

The effect of ammonium on protein processing and secretion from mouse insulinoma beta TC3 cells and recombinant mouse pituitary AtT-20 cells was investigated. Protein stored in granules of the regulated secretion pathway was discharged from cells with secretagogues, and the addition of new protein was encouraged by recharging cells in serum-containing medium with or without added ammonium. Ammonium at 6 mM inhibits the addition of insulin-related peptides to intracellular stores in both cell lines; it is not clear whether this effect is dose dependent for concentrations between 0 and 6 mM. There is a slight increase in insulin-related proteins secreted during recharging of cells in the presence of ammonium. Using reverse-phase high-performance liquid chromatography to separate proinsulin from insulin, we found that extracts from beta TC3 cells recharged in the presence of 6 mM ammonium contain significantly less insulin than control cells recharged in the absence of ammonium.