- 2016 Grants
- 2015 Grants
- 2014 Grants
- 2013 Grants
- 2012 Grants
- 2011 Grants
- 2010 Grants
- 2009 Grants
- 2008 Grants
- 2007 Grants
- 2006 Grants
- 2005 Pilot Study Grants
- 2004 Pilot Study Grants
- 2003 Pilot Study Grants
- 2002 Pilot Study Grants - Grants 1-6
- 2001 Pilot Study Grants - Grants 1-26
- 2000 Pilot Study Grants - Grants 1-21
- 2000 Pilot Study Grants - Grants 22-43
- 1998-1999 Pilot Study Grants - Grants 1-22
- 1998-1999 Pilot Study Grants - Grants 23-47
- Where to Apply for a Grant
- Research Dollars at Work
Funding from The Parkinson Alliance helped to finance the following Parkinson's research. Grantees were selected by scientific review committees of participating organizations. Updates will be posted, when available.
Project Title: Induction of Axon Re-Growth in Dopamine Neurons: A New Approach to Neurorestoration in Parkinson’s Disease
Grant Awarded to: Robert E. Burke, MD
Objective: To develop gene therapy approaches to induction of axon re-growth in the treatment of PD
Background: All connections between brain cells are made by long, thin neural fibers called axons. Axons can be thought of as the ‘wiring’ of the brain. We have recently presented evidence that in PD, there is a much greater loss of axons than brain cells. We have also shown that adult dopamine neurons can be induced to re-grow their axons if they are treated by a new gene therapy that re-activates the processes that mediate normal axon growth during development.
Methods/Design: We will modify genes that normally mediate axon growth during development to make them more potent. They will then be transferred to dopamine neurons using an AAV viral vector that. This approach will be tested in current animal models of PD.
Relevance to Parkinson’s disease: At the present time there is no existing therapy for PD that restores the axon connections between brain cells that have degenerated due to the disease. It is the loss of these axons that is responsible for the actual symptoms of the disease. Neuroscientists have long believed that in the adult brain, such axon connections, after they have been lost, can never be restored. We have shown that this traditional belief is not true; axons can be induced to re-grow and be restored. This new approach to treatment of PD offers new hope.
November 2012 Project Update:
We have previously shown that, contrary to the long-held belief that the adult central nervous system is incapable of long-range axon re-growth, lesioned dopamine neurons can be induced to grow new axons when they are transduced by adeno-associated viral (AAV) vector to express highly active forms of the kinase Akt. This molecule is known to play a role in the growth of axons during development. Our work has shown that even in maturity dopamine neurons remain responsive to it and generate new axons. In order to investigate the promise of this approach for the treatment of PD patients, we sought to determine how long after axon destruction the dopaminergic neurons remain capable of growing new axons. In order to more accurately model chronic human PD, we assessed the axon growth response following AAV treatment at six weeks following lesion. In the rodent lesion model, about 70% of dopamine neurons have degenerated at that time, so this would be comparable to PD of about ten years duration. We found in the lesion model that even after this long delay in AAV treatment there was a significant axon growth response (published in Molecular Therapy, 2012).
In order to make this approach safe for clinical use in the treatment of PD patients, we need to be able to control where the genes are expressed. Ideally it would be best to limit expression to the dopamine neurons where they are needed. One approach to achieve this specificity in cellular expression is to put the gene that induces axon growth under the control of a promoter that is expressed exclusively in dopamine neurons. One such promoter is the tyrosine hydroxylase (TH) promoter, which is active exclusively in catecholaminergic neurons such as dopamine neurons. We have successfully cloned three human TH promoter constructs and we are currently evaluating them for strength and specificity of expression.
Project Title: Predictors and Pathology of Speech decline in Parkinson’s disease
Grant Awarded to: Cindy Zadikoff, MD, MS, Colum MacKinnon, PhD, and Kristin Larsen, MA, CCC-SLP
1. Understand and quantify the effect of frequency (rate/Hz) on speech impairment in Parkinson’s disease
2. Quantify the effect of DBS on speech in patients with Parkinson’s disease as well as determine pre operative predictors of speech decline post operatively.
Background: Speech decline is a common early and progressive feature in Parkinson’s disease that has a variable response to medication. Previous work on finger tapping by Dr MacKinnon has shown a rate dependent impairment in repetitive movements (2 Hz barrier) . We would like to extend this theory to speech impairment in PD by using quantitative methods to study speech in PD.
We also know that STN DBS is considered a mainstay of therapy in patients with advanced Parkinson’s disease however speech can be worsened by stimulation as well as disease progression. The body of literature on speech, as affected by DBS, is sparse, but growing. However, there is still no good data on how to predict who will develop speech dysfunction in relation to DBS.
Methods/Design: Open label study to be conducted at Northwestern University Parkinson’s disease and Movement Disorders Center and Feinberg School of Medicine Dept of Rehab. in which consecutive subjects undergoing DBS for Parkinson’s disease are evaluated pre operatively with quantitative speech (and swallowing) analysis and again at 3 months post operatively when stimulation parameters are optimized from a motor standpoint. Pre operatively subjects are studied both ON and OFF meds. Post operatively subjects will be studied both ON STIM and OFF STIM/ON MEDS and OFF MEDS. Speech parameters will be analyzed and quantified in the above mentioned states and then predictors of decline will be ascertained using a multivariate model.
Relevance to Parkinson’s disease: Speech impairment is an early and progressive feature of Parkinson’s disease that has a variable response to medications and can lead to a significantly impaired quality of life. Understanding whether a frequency barrier exists in Parkinson’s disease may help explain why certain therapies work better than others and would help guide therapy for patients. Moreover, while DBS is a mainstay of therapy it is important for patients undergoing this therapy to understand the potential risks. Being able to predict and quantify who might be at risk for speech impairment in association with DBS would allow physicians to better advice and counsel patients re: the risks and benefits of undergoing deep brain stimulation and thus would allow patients to make more informed decisions.
November 2012 Project Update:
There are two distinct speech in DBS studies underway that are utilizing the funding provided by Parkinson Alliance.
The first involves a project in collaboration with Dr Colum Mackinnon.
Dr. MacKinnon has now moved to the Department of Neurology at the University of Minnesota. Nonetheless, we are continuing to recruit and test patients with STN-DBS. Dr. MacKinnon has set up a speech testing station in the Movement Disorders Laboratory at the University of Minnesota. The station includes a KayPentax Computerized Speech Lab (CSL) and software for the collection and analysis of speech and voice using standardized protocols. Dependent variables will include measures of rate, intensity and spectral analysis. In addition, we have interfaced the Speech system with our existing 3-D kinematic, multi-channel EMG and high-resolution EEG systems. This state-of-the-art set-up will allow us to quantify movement and muscle activity of the face and mouth during phonation and to examine cortical activity associated with normal and disordered speech. The DBS protocol will examine the changes in speech, muscle activity and EEG between the OFF-stimulation (OFF-meds) and ON-stimulation states. We have also initiated a collaboration with Dr. Matthew Johnson, Department of Biomedical Engineering, University of Minnesota, to examine the placement of the DBS macroelectrodes, volume of tissue activated, and its relationship to speech impairment during stimulation.
The second portion of the study involves clinical speech assessments in patients pre and post DBS in an effort to better clinically categorize the nature of the changes that occur s/p DBS and potentially identify pre operative predictors of speech decline to better counsel patients about risk of speech impairments s/p DBS.
12 patients have been seen for both a pre-op and a post-op visit. Of these, 11 had bilateral surgery and 1 had unilateral surgery.
18 subjects were referred for pre-operative assessment but were never referred back for post-operative evaluations. Of these, several did not end up having surgery (so there was no post-operative data to be obtained) or had complications following surgery and so were not referred for post-operative testing. Some of these were missed referrals. Unfortunately these are incomplete data sets and therefore cannot be included in the analysis.
The transcriptions and speech analysis on 5 of the 12 subjects with pre- and post- operative data have been completed thus far with the aim to complete the remaining analyses over the next few months. At that point we will plan to prepare and submit a manuscript for publication.
Project Title: Instrumented Balance and Gait to Improve Mobility Outcomes for DBS
Investigators: Fay B. Horak, PhD, Arash Salarian, PhD, Rajal Cohen, PhD, Matt Brodsky, MD,Oregon Health and Science University, Dept. of Neurology, Movement Disorders
Objectives: To improve mobility outcomes from DBS intervention with portable, objective measures of balance and gait before and after surgery.
Background: Clinical judgment and patient impressions are all that are currently used to determine the effectiveness of DBS surgery and stimulation optimization on balance and gait. The lack of a sensitive, objective measure of balance and gait may be limiting the effectiveness of DBS intervention. We recently showed that after 6 months of stimulation, many aspects of balance control are worse than before surgery but we don’t know if this worsening is due to the surgery procedure (before stimulation is started) and/or to poor programming of the stimulators.
Methods/Design: Dr. Horak’s laboratory recently developed a novel balance and gait system (Mobility Lab) using body worn sensors that allows therapists to obtain objective, finely graded measures of balance and gait in the same time it takes to do a subjective evaluation. We will use Mobility Lab to measure postural sway, step initiation, gait and turning before, and mediately after DBS surgery, before stimulators are turned on.
Thirty PD patients with stimulators placed bilaterally in STN or GPi will be tested. We hypothesize that objective measures of balance and gait will:
1) Provide an objective comparison of patients’ balance and gait before, compared to after, surgery as a basis for judging readiness for discharge;
2) Determine whether DBS electrode placement in GPi versus STN result in different immediate effects on balance and gait; and
3) Improve patient satisfaction with DBS.
Relevance to Parkinson’s Disease: This study will provide clinicians and patients better measures of balance and gait to make judgments about the effects of surgery and how to best tune their stimulators to improve mobility.: To improve mobility outcomes from DBS intervention with portable, objective measures of balance and gait before and after surgery.
November 2012 Project Update:
1. Objective balance and gait measures will provide objective comparison of a patient’s balance and gait before and day after surgery, as a basis to judge readiness for discharge.
2. Objective balance and gait measures will determine if GPi vs STN electrode placement results in different immediate effects on balance and gait.
3. Objective balance and gait measures will improve patient satisfaction with DBS.
Feasibility of Clinic Testing. Progress in the first year has been excellent. We have made the transition from testing subjects in the laboratory to during regular clinic appointments. The patients with Parkinson’s disease, who are candidates for DBS (deep brain stimulation) surgery, are seen by multiple providers in different clinical areas of the hospital before and after their surgery. An outpatient physical therapist in Rehabilitation makes a pre-surgery assessment. After surgery in-patient physical therapist in the hospital assesses their readiness to go home. One month after DBS implant patients are seen in the Neurology clinic by a physician’s assistant or neurologist to have their initial (and subsequent) programming visit. Finally, patients return to Rehabilitation services for a 3-6 month post-surgery physical therapy assessment.
We have successfully trained physical therapists and a physician’s assistant in the use of the Mobility Lab system (5 sensors, charging station, access point and laptop). We have also been able to incorporate use of the system during the subjects’ regularly scheduled clinical appointments and in hospital after surgery. The system is on a rolling cart that is easy to bring to the subject. The sensors are taken on and of subject and tests completed within 5 – 10 minutes. Space in the clinic can be restricted, compared to laboratory testing, since there are staff and patients moving about in clinic hallways and in hospital. However, we have been successful in finding 7m walking areas to test subjects. The system has a remote control so the tester can walk next to the subject for safely and still start/stop trials.
Subjects recruited. We have tested 35 subjects thus far and 16 subjects have had the DBS surgery (see flowchart in Figure 2). Eleven more subjects are waiting to be scheduled for surgery/verify that they qualify. The day after surgery we have tested 14 subjects, although one did not feel well enough to stand for testing. We have tested 8 subjects at the Day 30, programming appointment and 7 at the Day 90 follow-up PT appointment.
Data Analysis. We are in the process of data analysis and will report a complete summary across time-points after data collection is completed.
We collect the instrumented stand and walk test (iSAW). This test generates standing sway, step initiation, gait and turning measurements. Recently, we also added a newly developed inertial sensor test – instrumented compensatory stepping – to evaluate reactive balance. We previously found that reactive balance (regaining balance after a push or slip) gets worse in subjects with STN DBS, so it will be valuable to have a clinical test of this balance parameter.
Changes in postural sway during standing are evident when comparing subject data across visits. For instance, in one subject the sway area was abnormally large before surgery compared to healthy control subjects, but after the initial programming the stimulation the sway area was smaller.
We found that it is also possible to test subjects several times during optimization visits, and provide immediate summary reports, that the programming neurologist or physician’s assistant could use to aid in setting optimal stimulation parameters. Standard procedure has been for the clinician to simply watch the patient walk in the hallway after programming changes to judge gait, so this provides a more accurate way to quantify walking changes in response to different stimulator settings. In one subject for example – comparing off stimulation/medication to after the stimulation was turned on – the APA (anticipatory postural adjustment) associated with step initiation decreased in size (got worse) whereas the duration of turning improved. In contrast, the primary gait parameters were unchanged.
Project Title: Mechanisms of Pesticide Toxicity and their Role in Parkinson’s Disease
Grant Awarded to: Jeff Bronstein, MD, Ph.D and Marie-Francoise Chesselet, MD, Ph.D
Background: We have made significant progress in finding the causes of Parkinson’s disease (PD). Extensive genetic studies has revealed that a small number of patients have a strong genetic risk (approximately 5%) but for the majority of people with PD, genetic influences appear to contribute only a small portion of the risk. It is likely that environmental toxins and interactions between genes and the environment play the major role in causing PD.
Objective/Methods: Pesticide exposure has been suspected for several years to increase the risk of developing PD but only recently have we and others confirmed this association using objective measures. Importantly, we have identified a number of specific pesticides that increase the risk of PD. Our labs have been determining the mechanisms of pesticide toxicity. We have found that many pesticides inhibit 2 important processes believed to be involved in the pathogenesis of PD; proteasome and aldehyde dehydrogenase (ALDH) activity. The assays used to measure these activities require a piece of equipment (a multilabel plate reader) that can measure fluorescence and absorbance in a multiwell format. The plate reader we had been using to determine pesticide affects on proteasome and ALDH activity is old and recently broke, and we propose to use funds from the Parkinson Alliance/Team Parkinson to purchase a new one. It will be used by both of our labs to complete our screen for proteasome and ALDH inhibitors and determine their effects in cellular and animal models.
Relevance to Parkinson’s disease: Data from these studies will help determine how environmental toxins cause PD, which is an essential step towards risk factor reduction (i.e. disease prevention) and targeting new therapies to stop disease progression.
November 2012 Project Update:
Results: Our overall strategy to identify pesticides that might cause PD has been to screen a large library of pesticides for their ability to interfere with processes thought to be involved in the pathogenesis of PD (e.g. UPS and ALDH inhibition). Once identified, candidate toxins are then tested to see if exposure in people is associated with an altered risk of PD. Assays for ALDH and UPS activity were developed to determine enzyme activity in microtitier plates and the library of toxins was screened.
We identified several pesticides that inhibit ALDH and UPS activity at relevant concentrations. Benomyl is a fungicide that was studied further as a lead compound. Its metabolites were found to be responsible for its ALDH inhibitory activity and selectivity killed dopaminergic cells in both primary neuronal cultures and in zebrafish. Exposure to benomyl was found to be associated with a 200% increase in the risk of developing PD. When grouped together, 6 pesticides that inhibited ALDH were associated with even a greater risk of PD. Similar results were found with UPS inhibitors.
Conclusion/Relevance to Parkinson’s disease: Pesticides that inhibit ALDH and UPS increase the risk of developing PD. Other
environmental toxins and genetic variations may work through similar mechanisms and may account for a larger proportion of PD
Project Title: Understanding the role of the Frontal Cortex in Gait and Balance impairment in PD
Investigators/Authors: Beth Fisher PhD, PT; Ya-Yun Lee, MS, PT, Giselle M. Petzinger, MD, University of Southern California
Objective: To determine the role of the frontal cortex as a critical circuit involved in motor and cognitive problems in individuals with Parkinson’s Disease.
Background: The basal ganglia and the frontal cortex are important brain circuits responsible for key cognitive functions including the learning and maintenance of new motor skills. Evidence now exists that the loss of dopamine in PD leads to changes in frontal cortex circuitry to the basal ganglia resulting in motor difficulties including walking and balance. As a consequence of this degenerative circuit is that the learned motor skill, such as gait and balance, may only be performed accurately when the same environmental cues are present, but not when new environmental cues are delivered. For example, an individual with PD might demonstrate skillful ability to walk in the physical therapy clinic where they are practicing walking but then demonstrate difficulty with the same task when performed in an unfamiliar environment or outside of the clinic such as in a crowded shopping center or when their attention is drawn to other aspects of the environment. This dysfunction has been referred to as context-dependent motor learning. Progression of disease can lead to dramatic deficits in this circuitry such that patients experience severe and debilitating freezing of movement. We have observed abnormal over-activation of a region of the frontal cortex, called the dorsolateral prefrontal cortex (DLPFC). If this over-activation of the frontal cortex could be reversed then the consequences of this abnormal circuitry could represent a novel treatment for motor deficits in PD. Therefore, the purpose of this study is to test whether decreasing the activity of the DLPFC using the non-invasive technology of repetitive Transcranial Magnetic Stimulation ( rTMS) in individuals with PD would improve motor learning and function.
Methods/Design: We have developed a computer based task-specific paradigm that enables us to detect context-dependent motor learning in PD. We will use rTMS to temporarily suppress the cortical excitability of the DLPFC in people with PD. rTMS is a non-invasive technique that is able to directly stimulate regions of the brain such that they can be activated or suppressed. MRI is used to accurately locate the DLPFC in our subjects. By applying rTMS in individuals with PD, we will be able to determine whether context-dependent motor learning in PD is due to an over-activation of the DLPFC. We expect that decreasing the excitability of the DLPFC in PD, will results in a decrease in context-dependency learning in PD implicating the importance of this cortical brain region in PD symptomology.
Relevance to Parkinson’s disease: Ongoing studies in our lab examining features of abnormal cortical processing in individuals with PD has shown that there are severe deficits in learning especially in a specific type of learning termed context-dependent learning. This deficit may account for decline in gait, loss of balance, and eventually motor freezing especially when changing directions, starting and stopping, or entering new environments. Surprisingly, we have shown a deficit in this critical feature of motor control in patients at all stages of disease, and importantly, most patients are completely unaware that they have any kind of deficit. Therefore, these studies are critical because they will reveal to us important circuits within the brain that are responsible for some of the most difficult to treat clinical features of PD for which dopamine replacement therapy fails to combat. In addition, our findings will begin to identify critical regions of the brain that should be targeted for therapeutic treatment. We will be able to optimize treatments, both pharmacological and physical therapy based, to re-train the brain such that the learning of motor skills and preservation of motor memory can occur. For optimal treatment of the individual with PD, it is essential for clinicians to find training strategies that will maximize the ability to perform motor activities at a high level of skill regardless of the environment. It is our hope that such treatments will protect circuits within the brain that are lost as disease progresses and may in fact reverse potentially debilitating features of PD.
Project Title: Enhancing Repair in the Parkinsonian Brain
Investigators/Authors: Giselle M. Petzinger, MD; Michael W. Jakowec, PhD; Beth E. Fisher, PT/PhD, University of Southern California
Objective: To determine if combination therapy of intensive exercise and neurotrophic drugs can enhance brain repair in PD.
Background: Studies form our labs have shown that exercise in the form of intensive treadmill running leads to marked improvement in the connection and function between neurons of the basal ganglia responsible for normal movement. Specifically we observe an increase in the number of connections within the basal ganglia of parkinsonian (MPTP) mice after exercise and an improvement in the communication and function of these connections. We believe that it is the enhanced connection and function of neurons within the basal ganglia that are important in helping the repair processes needed to improve mobility and quality of life in PD. We hypothesize that one reason these connections are formed and their communication improves after exercise is through the synthesis of new proteins or building blocks of the brain that (i) help stabilize connections and (ii) help improve the signaling between these connections. Protein synthesis can often be enhanced through neurotrophic related compounds or drugs that help the cell function better. We hypothesize that combining exercise with a neurotrophic related compound will enhance brain recovery by facilitating the synthesis of proteins important in making connections and establishing signaling.
Methods/Design: For these studies, we will use a drug developed here at USC called ICG-001 that activates biochemical pathways that promote cell signaling and connections. ICG-001 will be delivered by mini-pumps into the brain of mice 5 days after they have been made parkinsonian through exposure to the dopamine-depleting drug MPTP. A group of mice will undergo treadmill running and will be compared with mice that receive drug but are sedentary and other mice that receive neither drug nor exercise. At the completion of the exercise all mice will be examined for the number of brain connections and the signaling of their connections (using electrophysiological techniques).
Relevance to Parkinson’s disease: Our studies support the role of exercise in promoting repair processes in PD. These processes include enhancing the connections between neurons within the basal ganglia, that are known to be impaired or lost in PD. We believe that results gained from this study can be translated into clinical studies in Parkinson’s disease that are focused on combining both exercise and neurotrophic related compounds that can be used to enhance brain function through facilitating the basal ganglia circuitry and its function.
Project Title: Developing screening procedures for pre-motor Parkinson’s disease
Investigators/Authors: J. William Langston, Parkinson’s Institute, Sunnyvale, CA
Objective: There is a major effort in the medical and scientific community to find ways to identify pre-motor PD. This will be particularly important to allow us to maximize the possibilities for success in clinical trials aimed at disease modification, as it would allow us to initiate such trials much earlier at a time when there is much less damage to the nervous system and there may be a much greater probability of success. Therefore, the goal of this work is to identify tools that can be used to screen for PD in the general population well before the potentially disabling motor or cognitive changes develop. If successful, this work could change the paradigm of diagnosis and treatment of PD.
Background: There is now abundance evidence that a substantial number of non-motor symptoms can precede the motor symptoms of PD, sometimes for more than a decade or more. This is increasingly referred to as pre-motor PD, and includes such symptoms and physical signs as loss of sense of smell, constipation, agitated and even violent dream-enacting behavior during sleep (known as REM-sleep behavior disorder or RBD), anxiety, depression, subtle changes in color vision, and a loss of normal heart rate variability. We now have very strong preliminary evidence that heart rate variability is decrease in pre-motor PD because the nerve fibers that are responsible normal variation in heart rate are lost in PD.
Methods/Design: In this study we will periodically examine individuals with signs of pre-motor PD (e.g. RBD or severe loss of smell) to see if decreased heart rate variability as measured by a simple EKG can identify individuals (1) with abnormal brain scans that suggest pre-motor PD and/or (2) who go on to actually develop PD.
Relevance to Parkinson’s disease: The short term goal of this research is to develop screening tools to identify pre-motor PD. If this can be accomplished, then populations of patients with pre-motor PD can participate in clinical trials aimed at disease modification at the earliest possible time in the course of their disease. Our long term goal is to develop a test battery for pre-motor PD that could be carried out in primary care physician’s office as part of an annual physical examination in all individuals over the age of 50 years – this would allow for annual screening of the general population when they are at increasing risk for PD. For example, data from a routine annual EKG to determine heart rate variability, and a simple scratch and sniff test for smell, could easily be incorporated into an annual physical examination and provide clues to pre-motor PD. Once such a screening battery has been developed and validated, patients with high risk can then be referred for more sophisticated imaging procedures and a full neurological evaluation. If and when current or future trials aimed at disease modification are successful, early intervention could lead to secondary disease prevention.
November 2012 Project Update:
Our prior work found strong preliminary evidence that heart rate variability determined from a simple 5-minute EKG is decreased in patients with REM-sleep behavior disorder (RBD). This project extends this work with a goal to develop sensitive and specific methods to identify prodromal (i.e., pre-motor) PD using non-invasive tests. We are evaluating two groups of study participants: 1) additional individuals with RBD, and 2) individuals with impaired olfactory function. In both study groups, we are analyzing EKGs to determine heart rate variability. In addition, we are combining the heart rate variability data with other clinical measures thought to be predictive of prodromal PD, such as constipation and subtle signs of motor impairment. Although most people with constipation or olfactory impairment do not have early PD, we are developing models that include multiple signs and symptoms in order to more accurately predict those at greatest risk. Some study subjects are undergoing brain SPECT imaging in order to assess whether they have reduced dopamine function in the part of the brain most affected in PD.
Work to date is very encouraging. We have analyzed more than 230 EKGs for heart rate variability. Some subjects have had multiple EKGs, so that we are able to look for changes over time. We continue to observe reduced heart rate variability in individuals with RBD, and we have recently found that heart rate variability may also be reduced in individuals with impaired olfactory function. We are currently developing algorithms that combine multiple non-specific signs and symptoms in order to accurately predict which persons have reduced dopamine function on brain scans. We will continue to follow study subjects prospectively to determine if those who we predict to be at highest risk eventually develop PD. If successful, we are hopeful that this work could markedly speed the development of therapeutic interventions to slow or prevent PD.
Click for a printer friendly version